Multiwfn official website: //www.umsyar.com/multiwfn. Multiwfn forum in Chinese: http://bbs.keinsci.com/wfn
You are not logged in.
Hello,
I do need to calculate the molecular electrostatic potentials for different aromatics at nucleus position. Therefor, at first I need to generate the .wfn file as I read. How exactly is the right order to do that, as Iam new to computational chemistry. Can I draw a molecule in WebMO, then somehow export that file to what? ORCA? Where calculation begins?
Thanks in Advance!
Offline
I am not familiar with webMO. I prefer to build initial molecular structure using GaussView, but it is a commerical software. Some freely available visualization softwares can also be used to bulid molecular 3D structure, such as Avogadro, gabedit... After finishing building the initial structure, export it to a common structure format, such as xyz, mol, mol2, pdb.
Multiwfn has a very convenient function for creating ORCA input file. Boot up Multiwfn and load the structure file, then inputoiin main menu, then in the newly appeared interface you can select a level to create corresponding input file. For geometry optimization and wavefunction analysis purpose, RI-B3LYP-D3(BJ)/def2-TZVP is suitable for most cases.
After generating the ORCA input file, run it by ORCA, you will obtain .gbw file, which contains wavefunction information (if the task is geometry optimization, it contains wavefunction of the final structure) and can be converted to .molden file, see beginning of Chapter 4 of Multiwfn manual for detail. The .molden file can be used as input file of Multiwfn to perform various kinds of analyses, including ESP analysis.
I have a very short video tutorial to illustrate how to use ORCA to perform a few simple tasks, which should be very useful for you:
Study geometry, vibration, IR spectrum and orbitals based on ORCA program and other codes
https://youtu.be/tiTmTbtbtig
Offline
Thanks for your fast reply!
I did finally manage to create the .molden file with ORCA.
Now, Iam interested in gaining the electrostatic potential at nucleus position. How do I get those values? I'd like to verify my results calculated with another software.
Which values from Multiwfn are comparable to those gained from Envision by Entos? They state the displayed values are "Electrostatic potential at nuclear centers".
Thank you in advance!
Last edited by Dialin (2022-04-29 10:43:38)
Offline
For example, boot up Multiwfn and input:
examples\H2O2.fch
1
a2
From screen you can find several quantities related to ESP at nuclear position of atom 2 (hydrogen in this case)
ESP from nuclear charges: 0.1000000000E+04 ESP from electrons: -0.7947626392E+01 Total ESP: 0.9920523736E+03 a.u. ( 0.2699512E+05 eV, 0.6225228E+06 kcal/mol) Total ESP without contribution from nuclear charge of atom 2: -0.9720067182E+00 a.u. ( -0.2644965E+02 eV, -0.6099439E+03 kcal/mol)
Usually "Total ESP without contribution from nuclear charge of atom 2" is of interest, it is ESP at nuclear position of H2, but the contribution from this nucleus is ignored. This value is closely related to pKa of the H2.
You can similarly calculate the value for other nuclei.
Offline
Thank you very much for this example.
I did exactly this but I got only negative ESP values. I expected to get also positive ones from those calculations as I did with the GFN1-xTB method (see screen above).
I'd like to evaluate the polar/non-polar character of small aromatic drug molecules and the sum of negative values gave me the correlation I was looking for (from GFN1-xTB method). Iam a bit curios right now. Maybe I have to modify the isovalue?
Thank you!
Offline
I don't know how the values in your map are calculated, according my knowledge, they are unlikely meaningful. Please do not insist on reproducing certain data (especially when the definition of the data is even unclear), but try to calculate really meaningful data via correct way.
For small drug molecule, there is no reason to use GFN1-xTB. Please bear in mind that GFN1-xTB is a fairly crude theory and the quality of resulting wavefunction is much poorer than popular DFT method (GFN-xTB can be viewed as a semi-empirical variant of DFT). To calculate molecular descriptors based on wavefunction, in most case DFT should be used, and B3LYP (the most popular DFT functional) is usually the preferred choice.
Offline
Thank you! I guess I have to switch over to Multiwfn program.
After I had a look at all ESP at nuclear position for two molecules, it turned out that those values were almost identical for the isomers of 1,2- and 1,4-Dihydroxybenzene:
1,2-DHB 1,4-DHB
a1 -14.77 -14.71
a2 -14.77 -14.76
a3 -14.72 -14.76
a4 -14.72 -14.71
a5 -14.77 -14.76
a6 -14.77 -14.76
a7 -1.1 -1.1
a8 -1.09 -1.09
a9 -1.09 -1.09
a10 -1.1 -1.09
a11 -22.35 -22.34
a12 -0.99 -0.99
a13 -22.34 -22.34
a14 -0.98 -0.98
This surprises me a bit, as I expect the electron density at the aromatic carbons to be way more different because of electron donating groups. Is there another property to look for with Multiwfn?
Offline
You should keep more decimal places, otherwise the difference cannot be distinguished.
In fact I don't know why you are interested in ESP at nuclear positions. If you hope to study number of electrons carried by various atoms under different chemical environments, usually atomic charge (also known as partial charge) is the quantity you should focus on, see Section 4.7 of Multiwfn manual for example on how to calculate it. Compared to ESP at nuclei, the value of atomic charge is much more sensitive to chemical environment and better reflects the status of atom.
Offline
Hi there,
I'm trying to convert the orca .gbw file to a .molden file but get the error:
Note: Albeit "orca_2mklpath" parameter in settings.ini has been defined, the orca_2mkl executable file cannot be located, therefore the .gbw file cannot be directly opened by Multiwfn
in the terminal i typed: 'which orca' and got '/home/epstein/orca/orca_5_0_3_linux_x86-64_openmpi411_part1/orca'
I then went into the settings.ini file and placed this path in the orcapath and orca_2mklpath locations yet am still receiving the same error message. What am I doing wrong here?
Thanks!
Offline
Hi there,
I'm trying to convert the orca .gbw file to a .molden file but get the error:
Note: Albeit "orca_2mklpath" parameter in settings.ini has been defined, the orca_2mkl executable file cannot be located, therefore the .gbw file cannot be directly opened by Multiwfnin the terminal i typed: 'which orca' and got '/home/epstein/orca/orca_5_0_3_linux_x86-64_openmpi411_part1/orca'
I then went into the settings.ini file and placed this path in the orcapath and orca_2mklpath locations yet am still receiving the same error message. What am I doing wrong here?
Thanks!
If your purpose is just converting .gbw to .molden, you do not need to use Multiwfn. Simply run orca_2mkl xxx -molden to convert xxx.gbw to xxx.molden.input.
orca_2mklpath in settings.ini of Multiwfn should be set only if you want to use Multiwfn to directly load .gbw file without manually converting it to .molden.input. In addition, orca_2mklpath should be set to the exact path (not folder) of executable file of orca_2mkl.
Offline
What is the outputted unit for those Mulliken-Charges? a.u.? To what is it related?
Thank you!
Offline