AMBER Archive (2008)

Subject: Re: AMBER: Electric field on atoms in equallibrium

From: Tom Darden (darden_at_niehs.nih.gov)
Date: Tue Apr 01 2008 - 14:26:32 CDT

i'm sure its in several textbooks---however just now i googled "pulay
forces"

the Hellmann-Feynman theorem states that the force on a nuclues for a
solution to a variational method (e.g. HF or DFT) is given by classical
electrostatics i.e. the electric field times the nuclear charge---thus at
equilibrium the field should be zero. However this assumes the
electronic structure does not depend implicitly on nuclear position.
An example would be plane wave basis sets that are independent of
nuclear positions, for which Hellmann-Feynman holds. However, for
typical Gaussian basis sets there are extra terms in the forces due to
the dependence of the representation of the wave function on the nuclear
position. These extra terms are termed the "Pulay forces" since P Pulay
characterized them some time ago. I think if you try to search on pulay
forces you may get to a decent explanation
hope this was useful

  On Tue, 1 Apr 2008, Lishan Yao wrote:

> Thank you for your explanation. It will be greatly appreciated if you can
> explain it in more details or suggest some references (books etc.).
>
> Best,
> Lishan
>
> ----- Original Message ----- From: "Tom Darden" <darden_at_niehs.nih.gov>
> To: <amber_at_scripps.edu>
> Sent: Tuesday, April 01, 2008 1:59 PM
> Subject: Re: AMBER: Electric field on atoms in equallibrium
>
>
>> if you use an atom-centric basis set such as in G03 the field will not be
>> zero--the reason is you are not correctly handling the change in field due
>> to movement of basis elements centered on the atom in question
>>
>>
>> On Tue, 1 Apr 2008, Lishan Yao wrote:
>>
>>> Hi Amber users:
>>> I am interested in electric field on atoms of a small peptide where
>>> electrons are treated explicitly. My question is that if I fully optimize
>>> the peptide by QM calculation (saying G03 program), will the field on each
>>> atom be zero?
>>>
>>> Thank you in advance!
>>>
>>> Best,
>>> Lishan
>>> -----------------------------------------------------------------------
>>> The AMBER Mail Reflector
>>> To post, send mail to amber_at_scripps.edu
>>> To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
>>>
>> -----------------------------------------------------------------------
>> The AMBER Mail Reflector
>> To post, send mail to amber_at_scripps.edu
>> To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
>>
>
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber_at_scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
>
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu