Molcas Forum

gnottoli

Member

Registered: 2025-11-10

Posts: 3

Issues with QM/MM dynamics calculation

Hello everyone,

I'm having trouble with ground state QM/MM dynamics calculations using the openmolcas/tinker interface (versions 23.02/6.3.3). The system is a solute in water solvent, with no link atoms, and most of the solvent molecules are a frozen shell around the moving ones (not introduced in the keyword file). After the first ~ 150 fs, one of the tip3p water molecules gets close to the QM solute and ends up collapsing into one of the carbon atoms, after which the whole system falls apart. I really don't understand how to prevent this, and thus I'm hoping someone might have insights into some specific keyword that should be added in the key file or others.
This is the input

> EXPORT MOLCAS_MOLDEN=ON

> COPY $CurrDir/$Project.xyz $WorkDir
> COPY $CurrDir/$Project.key $WorkDir
> COPY $CurrDir/OPLS_aa.prm $WorkDir
> COPY $CurrDir/$Project.velocity.xyz $WorkDir

 &GATEWAY
        Tinker
        basis = 6-31G*
        group = NoSym
        RICD

>>Foreach A in (1..200)

 &SEWARD
        DOANA

 &Espf
        External = Tinker

 &SCF
        charge=-1
 &ALASKA

 &Dynamix
        Velver
        DT = 41.3
        velo=1
        Ther=0

> COPY $WorkDir/$Project.xyz $CurrDir/$Project.Final.xyz
> COPY $WorkDir/$Project.velocity.xyz $CurrDir/$Project.newveloc.xyz
> End Do

And this is the .key file:

parameters OPLS_aa.prm
EXTRATERM NONE
QMMM  184
QM -1 34
MM -35 184
QMMM-MICROITERATION OFF
ACTIVE -1 184

The atom types for water are the standard tip3p (63 and 64) found in OpenMolcas/tinker/params/oplsaa.prm, whereas the ones for the solute are slightly modified (but I should note I encountered the same problem when using standard parameters for the solute as well). Any comment is welcome, thank you.