QMMM in cp2k

[Geng Sun]

Hello everyone 

    I want to study  supported metal cluster using molecular dynamics.

    Here, the substrate is represented by a slab which is metal oxide, and the metal cluster is put on the surface of the slab.

    I want to use the QMMM to reduce the computational cost by  using the force field to describe the interaction of  the bottom layers of slab model. The upper layers of slab

    and the metal atoms are left for DFT.

   Is that possible for CP2K to do the QMMM simulation in such case? If so, what is the most relevant parameters. 

   Thanks in advance !

   Geng Sun

    

[Teodoro Laino]

Hello,

as long as you have a sufficiently good force-field for the slab you can do that.

In your specific case, it should be enough to have parameters that capture most of the interaction within that material.

Did you search in literature for available classical force-field for your slab material ?

Teo

--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To post to this group, send email to cp...@googlegroups.com.
Visit this group at http://groups.google.com/group/cp2k.
For more options, visit https://groups.google.com/d/optout.

[Geng Sun]

Teo,

   The substrate for my model is Al2O3, which is ionic compound. This can be described by some classical force field.

   like

   http://dx.doi.org/10.1063/1.4832695

   or more general Tersoff  style potentials

   I am a newbie using QMMM in CP2K.  Two things confused me:

   1)  What kinds of classical force field could be used in CP2K,

   2)  What's the different of   CP2K_INPUT / FORCE_EVAL / MIXED,   and CP2K_INPUT / FORCE_EVAL / QMMM

        which one should I use ?

    Thanks again !

 Geng

在 2015年7月3日星期五 UTC+8下午2:47:40，Teo写道：

[Teodoro Laino]

On 03 Jul 2015, at 10:21, Geng Sun <sungen...@gmail.com> wrote:

Teo,

   The substrate for my model is Al2O3, which is ionic compound. This can be described by some classical force field.

   like

   http://dx.doi.org/10.1063/1.4832695

   or more general Tersoff  style potentials

   I am a newbie using QMMM in CP2K.  Two things confused me:

   1)  What kinds of classical force field could be used in CP2K,

You can use a multitude of classical potential. From third parties like AMBER and CHARMM to self-made potentials (including many-bodies, etc..). Have a look at the examples in the Fist regtest directory

   2)  What's the different of   CP2K_INPUT / FORCE_EVAL / MIXED,   and CP2K_INPUT / FORCE_EVAL / QMMM

        which one should I use ?

MIXED allows you to mix different Hamiltonians (total energy of your system is the sum of the terms of the different Hamiltonians). The best you can do is an ONIOM scheme (see literature).

If you want electronic coupling (between classical and quantum), you need to use the QMMM. Total energy here is still some of the classical + quantum but moreover you have also a QMMM term. So this is the one you should be using.

Teo