Question about xTB/DFT-MD and QM/MM in CP2K

[Hasan Tunçer]

Hello,

I wanted to ask whether an xTB/DFT-MD setup of is possible in CP2K. My idea is to treat my complex and nearby explicit waters with DFT, while describing the outer solvent region with xTB.

Related to this, would it be possible to perform a QM/MM simulation where the MM region is described by xTB? I am suspicious if the software is not okay with this approach.

If this is not a recommended approach, I would also appreciate guidance on the best practical way in CP2K.

Best regards,
Hasan

[Jürg Hutter]

Hi

the only way to do this is through an ONIOM like scheme using multiple FORCE_EVALs.
You can define an energy function of the type:

Etot[A+B] = ExTB[A+B] - ExTB[B] + EDFT[B]

with A and B are proper parts of your system.

Check online instructions and examples in the tests section.

regards
JH

________________________________________
From: cp...@googlegroups.com <cp...@googlegroups.com> on behalf of Hasan Tunçer <tuncer...@gmail.com>
Sent: Tuesday, March 24, 2026 8:34 PM
To: cp2k
Subject: [CP2K:22135] Question about xTB/DFT-MD and QM/MM in CP2K

--
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<mailto:cp2k+uns...@googlegroups.com>.
To view this discussion visit https://groups.google.com/d/msgid/cp2k/5e418c47-fa5c-41aa-8ee9-f09fd892dcc5n%40googlegroups.com<https://groups.google.com/d/msgid/cp2k/5e418c47-fa5c-41aa-8ee9-f09fd892dcc5n%40googlegroups.com?utm_medium=email&utm_source=footer>.