Hi Anthony
When people write 'hessian', they really mean mass weighted hessian.
That is the object you diagonalize to get normal modes, which you then
project onto bonds.
That mass weighting is what makes the force on an H and on a C the SAME
for H-C
is that ok ?
adrian
On 10/15/15 2:47 PM, Nash, Anthony wrote:
> Hi Dac,
>
> Many thanks for your input, I appreciate it.
>
> Completely agree with you, which is why I am extremely confused how a
> number (and not a small one) of publications have claimed to use second
> order derivation of cartesian coordinates (hessian) from QM calculations.
> An extensive example: "Systematic Derivation of AMBER Force Field
> Parameters
> Applicable to Zinc-Containing Systems². In their abstract:
>
> ³...In this study, we have derived bond stretching and angle-bending
> parameters applicable to zinc-containing systems which are compatible with
> the AMBER force field. A total of 18 model systems were used to mimic the
> common coordination configurations observed in the complexes formed by
> zinc-containing metalloproteins. The Hessian matrix of each model system
> computed at the B3LYP/6-311++G(2d,2p) level was then analyzed by
> Seminario¹s method to derive the desired force Constants.."
>
> Many thanks
> Anthony
>
>
>
> On 15/10/2015 17:35, "David A Case" <dacase.scarletmail.rutgers.edu> wrote:
>
>> On Thu, Oct 15, 2015, Nash, Anthony wrote:
>>> the force constant for A-B is not the same as B-A.
>>> So using second
>>> order tensors, the force constant for C-H is much greater than the
>>> force constant for H-C (where C is seen is the displacing atom, and H
>>> experiences the force).
>>>
>> This is not the way molecular mechanics potentials work. The bond energy
>> doesn't know anything about what is the "displacing atom". It just looks
>> at the distance from one atom to the next (which clearly doesn't depend
>> upon
>> which order is used.)
>>
>> If you wish to fit molecular mechanics potentials to quantum data, there
>> are
>> lots of tools that will help do that. But I don't know of any that start
>> from
>> the sort of non-symmetric tensors that you are thinking of.
>>
>> Maybe someone on the list will be able to chime in.
>>
>> ....dac
>>
>>
>> _______________________________________________
>> AMBER mailing list
>> AMBER.ambermd.org
>> http://lists.ambermd.org/mailman/listinfo/amber
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
--
Dr. Adrian E. Roitberg
Professor.
Department of Chemistry
University of Florida
roitberg.ufl.edu
352-392-6972
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Oct 15 2015 - 12:00:06 PDT
