odd DFT-BSSE energy output...

From NWChem

Viewed 2876 times, With a total of 6 Posts
Jump to: navigation, search

Gets Around
Threads 30
Posts 63
this link provides an example of calculating DFT energy HF dimer.
http://www.nwchem-sw.org/index.php/Top-level#TASK_Directive_for_BSSE_calculations

as:
title dimer
start dimer
geometry units angstrom
  symmetry c1 
  F 1.47189 2.47463 -0.00000 
  H 1.47206 3.29987 0.00000  
  F 1.46367 -0.45168 0.00000 
  H 1.45804 0.37497 -0.00000
end
basis "ao basis" 
  F library 6-31G 
  H library 6-31G 
  bqF library F 6-31G 
  bqH library H 6-31G
end
dft; xc slater 1.0 vwn_5 1.0; direct; end
bsse 
 mon first 1 2 
 mon second 3 4
end
task dft energy


But this results in an output like :
            BSSE error =       0.002710838142
  Supermolecular energy =    -199.468197200704
       Corrected energy =    -199.465486362562

Now,
1. is the Corrected-energy represented here in Hartee unit.?
2. If it is in Hartee, then how can a hydrogen bond be of ~199Hartee .?
3. else, I am missing something ..!
please comment.

thank you.
Edited On 4:36:53 AM PDT - Tue, May 27th 2014 by Neo

Gets Around
Threads 18
Posts 133
You are actually looking at the total energy for one of the HF monomer calculations included in the BSSE correction procedure.

As stated in the NWChem manual, there are 2N + 1 energy calculations involved in the correction procedure for a system of N monomers. In this case of the HF dimer that means 5 calculations total, and if you search the output for "Total DFT Energy" you will indeed see 5 matching lines. There is one calculation for the dimer, another calculation for the first monomer without ghost atoms, another calculation for the first monomer with ghost atoms, another calculation for the second monomer without ghost atoms, and finally a calculation for the second monomer with ghost atoms.

To find the actual BSSE-corrected dimer system energy look for the line starting "Corrected energy" near the end of the file.

None of the aforementioned energies are supposed to represent the strength of the hydrogen bond in the HF dimer. They all represent system energies in Hartree, where the starting point is separation of nuclei and all electrons in the system by infinite distance. Of course these energies are enormous compared to ordinary chemical reactions; no chemist starts his bench work with a cylinder of bare fluorine nuclei! To translate these absolute system energies to quantities of interest to chemists you need to calculate differences of absolute energies in related systems. Here, to estimate the hydrogen bond strength, you would want to take the difference in calculated energies between the HF dimer system and the sum of the two HF monomer systems in isolation. This very small residual is the energy that can be considered as hydrogen bonding. Especially with weak bonds like hydrogen bonds the basis set superposition error can be significant compared to the quantity of interest (here, the HF dimer hydrogen bond strength).

This example calculation from the manual produces the information necessary to estimate the hydrogen bond strength in HF dimer, including corrections for BSSE that could skew the result, but you must do the arithmetic of system sums and differences yourself.

Here is a detailed description of the BSSE correction, cited in the section of the NWChem manual you linked to: http://www.researchgate.net/publication/234965987_How_does_basis_set_superposition_error_c...

Gets Around
Threads 30
Posts 63
Quote:Mernst May 26th 5:09 pm
You are actually looking at the total energy for one of the HF monomer calculations included in the BSSE correction procedure.

As stated in the NWChem manual, there are 2N + 1 energy calculations involved in the correction procedure for a system of N monomers. In this case of the HF dimer that means 5 calculations total, and if you search the output for "Total DFT Energy" you will indeed see 5 matching lines. There is one calculation for the dimer, another calculation for the first monomer without ghost atoms, another calculation for the first monomer with ghost atoms, another calculation for the second monomer without ghost atoms, and finally a calculation for the second monomer with ghost atoms.

To find the actual BSSE-corrected dimer system energy look for the line starting "Corrected energy" near the end of the file.

None of the aforementioned energies are supposed to represent the strength of the hydrogen bond in the HF dimer. They all represent system energies in Hartree, where the starting point is separation of nuclei and all electrons in the system by infinite distance. Of course these energies are enormous compared to ordinary chemical reactions; no chemist starts his bench work with a cylinder of bare fluorine nuclei! To translate these absolute system energies to quantities of interest to chemists you need to calculate differences of absolute energies in related systems. Here, to estimate the hydrogen bond strength, you would want to take the difference in calculated energies between the HF dimer system and the sum of the two HF monomer systems in isolation. This very small residual is the energy that can be considered as hydrogen bonding. Especially with weak bonds like hydrogen bonds the basis set superposition error can be significant compared to the quantity of interest (here, the HF dimer hydrogen bond strength).

This example calculation from the manual produces the information necessary to estimate the hydrogen bond strength in HF dimer, including corrections for BSSE that could skew the result, but you must do the arithmetic of system sums and differences yourself.

Here is a detailed description of the BSSE correction, cited in the section of the NWChem manual you linked to: http://www.researchgate.net/publication/234965987_How_does_basis_set_superposition_error_c...



Dear Mernst,
Thank you for your suggestions. But I still have some points to clarify.
1. below i have pasted what i got as my BSSE corrected energy, when i ran the same code.
2. If it is in Hartee too then how should I interpret this huge energy..!?

            BSSE error =       0.002710838142
  Supermolecular energy =    -199.468197200704
       Corrected energy =    -199.465486362562


3. did i take wrong turning somewhere.?

thank you.

Gets Around
Threads 18
Posts 133
The corrected energy is the total energy of the HF dimer system after correcting for basis set superposition error. The supermolecular energy is the total energy of the HF dimer system before corrections for basis set superposition error. Both the corrected and uncorrected values are for the full energy of the HF dimer system, from a reference point of "complete separation of nuclei and electrons." If you want to calculate a quantity that can be interpreted as the hydrogen bond energy between the two HF molecules, you need to add up the energy of the two HF monomers calculated in isolation (this is present earlier in your output file, look for the lines that start Total DFT energy = -99...., use the monomer energies from the calculations that didn't include ghost atoms) and subtract that sum from the corrected full-system energy of 199.465486362562 Hartree. That gives a very small number that, converted from Hartree to kcal/mol, looks to be in the right ballpark for hydrogen bond strength.

There is more work to be done to determine if this sample procedure provides a general template you could use to calculate hydrogen bonding between pairs of molecules. Is the DFT approach used here generally and sufficiently accurate? Do you need experimental geometry as an input, or can you generate appropriate system geometries? These easy questions could take a staggering effort to answer without reading prior research. I presume that you have access to an academic library where you can perform a literature search on topics like quantum chemical modeling of hydrogen bonding. If you do not have access to a good library system, try running queries on Google Scholar. A fair number of famous/prominent papers on a specific subtopic will be hosted publicly somewhere and Google Scholar can guide you to them even if your library does not subscribe to the originating journal. That's how I found the paper on BSSE counterpoise correction that is cited by the NWChem manual. If you are trying to use an NWChem feature that you haven't used before, note what papers the software authors cited and track them down so you can understand better. Try to find related review articles as well so you can get a sense for the surrounding context of the methodology and research questions without trying to personally synthesize it from dozens of more focused papers.

Good luck with your research!
Edited On 1:21:44 PM PDT - Tue, May 27th 2014 by Mernst

Gets Around
Threads 30
Posts 63
Quote:Mernst May 27th 1:19 pm
The corrected energy is the total energy of the HF dimer system after correcting for basis set superposition error. The supermolecular energy is the total energy of the HF dimer system before corrections for basis set superposition error. Both the corrected and uncorrected values are for the full energy of the HF dimer system, from a reference point of "complete separation of nuclei and electrons." If you want to calculate a quantity that can be interpreted as the hydrogen bond energy between the two HF molecules, you need to add up the energy of the two HF monomers calculated in isolation (this is present earlier in your output file, look for the lines that start Total DFT energy = -99...., use the monomer energies from the calculations that didn't include ghost atoms) and subtract that sum from the corrected full-system energy of 199.465486362562 Hartree. That gives a very small number that, converted from Hartree to kcal/mol, looks to be in the right ballpark for hydrogen bond strength.

There is more work to be done to determine if this sample procedure provides a general template you could use to calculate hydrogen bonding between pairs of molecules. Is the DFT approach used here generally and sufficiently accurate? Do you need experimental geometry as an input, or can you generate appropriate system geometries? These easy questions could take a staggering effort to answer without reading prior research. I presume that you have access to an academic library where you can perform a literature search on topics like quantum chemical modeling of hydrogen bonding. If you do not have access to a good library system, try running queries on Google Scholar. A fair number of famous/prominent papers on a specific subtopic will be hosted publicly somewhere and Google Scholar can guide you to them even if your library does not subscribe to the originating journal. That's how I found the paper on BSSE counterpoise correction that is cited by the NWChem manual. If you are trying to use an NWChem feature that you haven't used before, note what papers the software authors cited and track them down so you can understand better. Try to find related review articles as well so you can get a sense for the surrounding context of the methodology and research questions without trying to personally synthesize it from dozens of more focused papers.

Good luck with your research!


I got a value of ~6kcal/mol for this HF_dimer_calculation.
and thank you for your suggestions.

Gets Around
Threads 30
Posts 63
Quote:Mernst May 27th 1:19 pm
The corrected energy is the total energy of the HF dimer system after correcting for basis set superposition error. The supermolecular energy is the total energy of the HF dimer system before corrections for basis set superposition error. Both the corrected and uncorrected values are for the full energy of the HF dimer system, from a reference point of "complete separation of nuclei and electrons." If you want to calculate a quantity that can be interpreted as the hydrogen bond energy between the two HF molecules, you need to add up the energy of the two HF monomers calculated in isolation (this is present earlier in your output file, look for the lines that start Total DFT energy = -99...., use the monomer energies from the calculations that didn't include ghost atoms) and subtract that sum from the corrected full-system energy of 199.465486362562 Hartree. That gives a very small number that, converted from Hartree to kcal/mol, looks to be in the right ballpark for hydrogen bond strength.

There is more work to be done to determine if this sample procedure provides a general template you could use to calculate hydrogen bonding between pairs of molecules. Is the DFT approach used here generally and sufficiently accurate? Do you need experimental geometry as an input, or can you generate appropriate system geometries? These easy questions could take a staggering effort to answer without reading prior research. I presume that you have access to an academic library where you can perform a literature search on topics like quantum chemical modeling of hydrogen bonding. If you do not have access to a good library system, try running queries on Google Scholar. A fair number of famous/prominent papers on a specific subtopic will be hosted publicly somewhere and Google Scholar can guide you to them even if your library does not subscribe to the originating journal. That's how I found the paper on BSSE counterpoise correction that is cited by the NWChem manual. If you are trying to use an NWChem feature that you haven't used before, note what papers the software authors cited and track them down so you can understand better. Try to find related review articles as well so you can get a sense for the surrounding context of the methodology and research questions without trying to personally synthesize it from dozens of more focused papers.

Good luck with your research!


I got a value of ~6kcal/mol for this HF_dimer_calculation.
and thank you for your suggestions.

Gets Around
Threads 30
Posts 63
Quote:Mernst May 27th 1:19 pm
The corrected energy is the total energy of the HF dimer system after correcting for basis set superposition error. The supermolecular energy is the total energy of the HF dimer system before corrections for basis set superposition error. Both the corrected and uncorrected values are for the full energy of the HF dimer system, from a reference point of "complete separation of nuclei and electrons." If you want to calculate a quantity that can be interpreted as the hydrogen bond energy between the two HF molecules, you need to add up the energy of the two HF monomers calculated in isolation (this is present earlier in your output file, look for the lines that start Total DFT energy = -99...., use the monomer energies from the calculations that didn't include ghost atoms) and subtract that sum from the corrected full-system energy of 199.465486362562 Hartree. That gives a very small number that, converted from Hartree to kcal/mol, looks to be in the right ballpark for hydrogen bond strength.

There is more work to be done to determine if this sample procedure provides a general template you could use to calculate hydrogen bonding between pairs of molecules. Is the DFT approach used here generally and sufficiently accurate? Do you need experimental geometry as an input, or can you generate appropriate system geometries? These easy questions could take a staggering effort to answer without reading prior research. I presume that you have access to an academic library where you can perform a literature search on topics like quantum chemical modeling of hydrogen bonding. If you do not have access to a good library system, try running queries on Google Scholar. A fair number of famous/prominent papers on a specific subtopic will be hosted publicly somewhere and Google Scholar can guide you to them even if your library does not subscribe to the originating journal. That's how I found the paper on BSSE counterpoise correction that is cited by the NWChem manual. If you are trying to use an NWChem feature that you haven't used before, note what papers the software authors cited and track them down so you can understand better. Try to find related review articles as well so you can get a sense for the surrounding context of the methodology and research questions without trying to personally synthesize it from dozens of more focused papers.

Good luck with your research!


I got a value of ~6kcal/mol for this HF_dimer_calculation.
and thank you for your suggestions.


Forum >> NWChem's corner >> Compiling NWChem



Who's here now Members 0 Guests 1 Bots/Crawler 0


AWC's: 2.5.10 MediaWiki - Stand Alone Forum Extension
Forum theme style by: AWC