[molpro-user] Question on calculation of electrostatic properties

[Werner Győrffy]

Dear Dave,

There are three way how one can compute expectation value is Molpro: by 
using EXPEC, PROPERTY, or MATROP. Please find your modified input below, 
where I gave an example for each case. Out of those, MATROP should be 
used for computing molecular electrostatic potential (MEP) for many points.

Furthermore, please consider the following

1,

One-electron properties can be computed  with or without taking into 
account the orbital relaxation (in analytical energy derivatives), which 
can be specified as 'expec,relax,...' or 'expec,norelax,...'. 
Furthermore, the density matrix can be saved by the DM card. For example,

{mp2;expec,relax,pot,,-1.00288,3.86820,0.00000;
dm,4300.2;}

2,

Molecular electrostatic potential has an electronic and a nuclear 
contribution. By using the PROPERTY program, these two contributions are 
automatically considered, this is however not the case in MATROP. After 
loading the density matrix and the integrals POT [should be defined here 
as POT(0)!]

load,den,den,4300.2;
load,pot,oper,pot(0),,-1.00288,3.86820, 0.001

The nuclear part must be added explicitly in the TRACE operation as

TRACE,variable, mat1, mat2,[factor],[facnuc]
Computes variable = factor*trace(mat1 
 mat2) + facnuc*opnuc.

In our example,

trace,ep,den,pot,1,1

3,

I recommend you to download latest patch to Molpro 2012.1 or the 
2012.1.nightly tarball from the Molpro website which is available to 
licensees. There are some recent bug fixes related to properties.

---

I hope it helps.

Regards,

Werner.


Your input modified:
---
memory,100,m

geometry={
   H1    0.9572000000000   0.0000000000000   0.0000000000000
   O2    0.0000000000000   0.0000000000000   0.0000000000000
   H3   -0.2399900000000   0.9266300000000   0.0000000000000
}

basis=cc-pvdz

{hf;}
! Compute MEP by using EXPEC
{mp2;expec,relax,pot,,-1.00288,3.86820,0.00000;
dm,4300.2;}

! Compute MEP by using PROPERTY
{property
density,4300.2
pot,,-1.00288,3.86820,0.00000
}

! Compute MEP by using MATROP
{matrop;
load,den,den,4300.2;
load,pot,oper,pot(0),,-1.00288,3.86820, 0.001
trace,ep,den,pot,1,1
load,pot,oper,pot(0),,-1.00288,3.86820, 0.002
trace,ep,den,pot,1,1
load,pot,oper,pot(0),,-1.00288,3.86820, 0.003
trace,ep,den,pot,1,1
...
...
...
load,pot,oper,pot(0),,-1.00288,3.86820, 0.249
trace,ep,den,pot,1,1
load,pot,oper,pot(0),,-1.00288,3.86820, 0.250
trace,ep,den,pot,1,1}

---

On 02/03/2015 06:59 PM, David Cerutti wrote:
> I hope this isn't a terribly pedestrian question, but is there a still a
> low limit to the number of points at which one can compute the
> electrostatic potential around a molecule with MolPro?  I found this as
> about the most helpful response, but it is for old code:
>
> http://www.molpro.net/pipermail/molpro-user/2008-September/002606.html
>
> I wasn't able to get very far with the example in that thread, but I was
> able to run the following input:
>
>  >>>
> memory,500,m
>
> ! AAA2
> geometry={
>    H1    0.9572000000000   0.0000000000000   0.0000000000000
>    O2    0.0000000000000   0.0000000000000   0.0000000000000
>    H3   -0.2399900000000   0.9266300000000   0.0000000000000
> }
> gdirect
> basis=cc-pvtz
> hf
> df-mp2
>
> !{matprop
> !read, savmat, DEN, charge, file=density.mat
> !save, savmat, 2140.2, DENSITY}
>
> {property
> density
> pot,,   4.00000,   0.00000,   0.00000
> pot,,  -1.00288,   3.86820,   0.00000
> pot,,   0.00000,   0.00000,   4.00000
> }
> <<<
>
> Once I figured out that geometry={ } gets specified in Angstroms and
> {property } coordinates get specified in Bohr I was in good shape.
> However, it seems that the limit of thirty points using this method is
> still in effect.  Is there a better way that I can compute the
> electrostatic potential for perhaps 5000 points following one MP2
> calculation?
>
> Thanks,
>
> Dave
>
>
>
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user
>