Partial charges outside of the atoms center
Gets Around
10:47:41 AM PDT - Fri, Sep 14th 2012
Sorry that I renew the question, but I really would like to find a way to exclude from the consideration points corresponding to the center of the atoms during the procedure of determination of the partial atomic charges.

I understand that I may include some additional points for the PDC procedure using X instead of the name of the element. But the program uses these new points and centers of atoms too. I need to the partial charges only on my points without consideration centers of atoms at all.
Could you suggest some sort of the trick?

Forum Vet
2:15:56 PM PDT - Tue, Sep 18th 2012
I'm sorry, but I am a little lost as to what you are trying to do, how you are trying to do it (whcih module in NWChem, input files would help).

Bert


Quote:P99 Sep 14th 5:47 pm
Sorry that I renew the question, but I really would like to find a way to exclude from the consideration points corresponding to the center of the atoms during the procedure of determination of the partial atomic charges.

I understand that I may include some additional points for the PDC procedure using X instead of the name of the element. But the program uses these new points and centers of atoms too. I need to the partial charges only on my points without consideration centers of atoms at all.
Could you suggest some sort of the trick?

Gets Around
2:30:29 AM PDT - Wed, Sep 19th 2012
For example, consider the molecule HF.
I can run the usual calculation:

start nw

geometry    
 F      0.0   0.0   0.0883587
 H      0.0   0.0  -0.8331587  
end 

basis spherical
 * library aug-cc-pvtz
end 

esp 
 recalculate 
 probe   0.07
 range   0.3
 factor  1
 spacing 0.02
end 

property 
 dipole 
 quadrupole 
end 

mp2 
 tight 
 freeze atomic 
end 

task MP2 gradient
set "esp:input vectors" nw.mp2nos
task esp


But I want to calculate a partial charge is not in the position coordinates of nucleus 
H      0.0   0.0  -0.8331587

but for example in the point 
X      0.0   0.0  -0.5


I can enter the new position for the calculation, but how to remove the old one?




Quote:Bert Sep 18th 2:15 pm
I'm sorry, but I am a little lost as to what you are trying to do, how you are trying to do it (whcih module in NWChem, input files would help).

Bert

Forum Vet
4:27:39 PM PDT - Mon, Sep 24th 2012
If I understand you correctly, what you are trying to do is to calculate the potential on a grid as a result of the (in the example) HF molecule and then find the partial charges at different points away from the atom centers, that still mimics the potential on that grid (which is the purpose of the ESP fitting code). I can see what youw ant to do used in for example replacing multiple atoms, each with partial charges, with one to do coarse-graining in MD.

Unfortunately, you would have to modify the ESP code to do this.

If you need the electrostatic potential at a point in space, the properties module can do that, see below.

Bert



start nw
geometry
F      0.0   0.0   0.0883587

H      0.0   0.0  -0.8331587
end
basis spherical
* library aug-cc-pvtz
end
esp
recalculate

probe   0.07

range   0.3

factor  1

spacing 0.02
end
property
dipole

quadrupole

esp

vectors nw.mp2nos
end
mp2
tight

freeze atomic
end
task MP2 gradient
task MP2 property
set "esp:input vectors" nw.mp2nos
task esp
geometry noautoz noautosym
F      0.0   0.0   0.0883587

H      0.0   0.0  -0.8331587

X      0.0   0.0  -0.5
end
task MP2 gradient
task MP2 property
set "esp:input vectors" nw.mp2nos
task esp


Quote:P99 Sep 19th 9:30 am
For example, consider the molecule HF.
I can run the usual calculation:

start nw

geometry    
 F      0.0   0.0   0.0883587
 H      0.0   0.0  -0.8331587  
end 

basis spherical
 * library aug-cc-pvtz
end 

esp 
 recalculate 
 probe   0.07
 range   0.3
 factor  1
 spacing 0.02
end 

property 
 dipole 
 quadrupole 
end 

mp2 
 tight 
 freeze atomic 
end 

task MP2 gradient
set "esp:input vectors" nw.mp2nos
task esp


But I want to calculate a partial charge is not in the position coordinates of nucleus 
H      0.0   0.0  -0.8331587

but for example in the point 
X      0.0   0.0  -0.5


I can enter the new position for the calculation, but how to remove the old one?




Quote:Bert Sep 18th 2:15 pm
I'm sorry, but I am a little lost as to what you are trying to do, how you are trying to do it (whcih module in NWChem, input files would help).

Bert

Gets Around
8:10:12 AM PDT - Wed, Sep 26th 2012
Quote:Bert Sep 24th 4:27 pm
If I understand you correctly, what you are trying to do is to calculate the potential on a grid as a result of the (in the example) HF molecule and then find the partial charges at different points away from the atom centers, that still mimics the potential on that grid (which is the purpose of the ESP fitting code).

Yes


Quote:Bert
I can see what youw ant to do used in for example replacing multiple atoms, each with partial charges, with one to do coarse-graining in MD.

No. No replacement.
I just want to change the position of the second point (H) during partial charges fitting.

Forum Vet
9:19:10 AM PDT - Wed, Sep 26th 2012
Simply change the geometry to what you want after the MP2 gradient calculation. Don't know what the results mean, but you can do it. See below for a couple of options.

Bert


start nw
geometry 
F      0.0   0.0   0.0883587

H      0.0   0.0  -0.8331587
end
basis spherical
* library aug-cc-pvtz
end
esp 
recalculate 

probe   0.07

range   0.3

factor  1

spacing 0.02
end
mp2 
tight 

freeze atomic
end
task MP2 gradient
set "esp:input vectors" nw.mp2nos
geometry 
F      0.0   0.0   0.0883587

H      0.0   0.0  -0.5000000
end
task esp
geometry 
F      0.0   0.0   0.0883587

H      0.0   0.0  -0.8331587

X      0.0   0.0  -0.5000000
end
task esp



Quote:P99 Sep 26th 3:10 pm
Quote:Bert Sep 24th 4:27 pm
If I understand you correctly, what you are trying to do is to calculate the potential on a grid as a result of the (in the example) HF molecule and then find the partial charges at different points away from the atom centers, that still mimics the potential on that grid (which is the purpose of the ESP fitting code).

Yes


Quote:Bert
I can see what youw ant to do used in for example replacing multiple atoms, each with partial charges, with one to do coarse-graining in MD.

No. No replacement.
I just want to change the position of the second point (H) during partial charges fitting.

Gets Around
6:15:15 AM PDT - Fri, Sep 28th 2012
Unfortunately, this is not working.

1) Problem is that the program automatically moves the coordinates.
I solved it with words “nocenter noautosym”.

2) For the first time “task esp” gives the correct dipole moment, but in the second case (with the new geometry) both dipole and partial charges are greatly reduced.
In fact, the dipole moment should remain near the same by increasing the partial charges.


start nw

geometry nocenter noautosym
 F      0.0   0.0   0.0883587
 H      0.0   0.0  -0.8331587  
end

basis spherical
 * library aug-cc-pvdz
end

esp
 recalculate 
 probe   0.07
 range   0.3
 factor  1
 spacing 0.02
end

mp2
 tight 
 freeze atomic 
end

task MP2 gradient
set "esp:input vectors" nw.mp2nos
task esp

geometry nocenter noautosym
 F      0.0   0.0   0.
 H      0.0   0.0  -0.8331587  
end

task esp

Forum Vet
12:46:03 PM PDT - Fri, Sep 28th 2012
I think to get what you really want will require you to modify the code.

Effectively, what ESP does it generates a grid and then looks at the nuclear and electronic contributions to the electrostatic potential at each point on the grid. Then it tries to fit charges on the atom position that represent this potential.

The trick I though of below effectively generates a electrostatic potential with the nuclear component from the atoms closer to each other plus the electronic contribution from the atoms further away.

What you will have to do in the code is to have the code generate the electrostatic potential on the grid using the original geometry, and then only in the fitting part adjust the atom positions to where you want to calculate the partial charges.

Still not clear what it will mean as the electron density assumes the atomic charge in one location and how you move that charge away.

Bert

Quote:P99 Sep 28th 1:15 pm
Unfortunately, this is not working.

1) Problem is that the program automatically moves the coordinates.
I solved it with words “nocenter noautosym”.

2) For the first time “task esp” gives the correct dipole moment, but in the second case (with the new geometry) both dipole and partial charges are greatly reduced.
In fact, the dipole moment should remain near the same by increasing the partial charges.


start nw

geometry nocenter noautosym
 F      0.0   0.0   0.0883587
 H      0.0   0.0  -0.8331587  
end

basis spherical
 * library aug-cc-pvdz
end

esp
 recalculate 
 probe   0.07
 range   0.3
 factor  1
 spacing 0.02
end

mp2
 tight 
 freeze atomic 
end

task MP2 gradient
set "esp:input vectors" nw.mp2nos
task esp

geometry nocenter noautosym
 F      0.0   0.0   0.
 H      0.0   0.0  -0.8331587  
end

task esp


Forum >> NWChem's corner >> General Topics