ONIOM in CP2K

[daniel Storm]

Hi everyone,

To provide some context, I am working with an iridium organometallic salt, [Ir(PONOP)(H)Me][BAr^F₄] and my goal here is to do a QM/MM type calculation where my MM region is treated with xTB.

I have tried my best to piece together an ONIOM input file from what I’ve seen on the internet but it’s not quite right. I keep getting a CPASSERT error suggesting that there is an issue with how the fragments are numbered? I’ve double-checked those and it looks correct to me.

My QM region (Fragment 1) are atoms 406-545

My xTB region (Fragment 2) are atoms 1-405

Any help with this issue or information on how to correctly construct an ONIOM input file would be much appreciated. I have attached my attempt and output files below.

Thanks,

Daniel

[Marcella Iannuzzi]

Hi 

I think you need to specify 

MULTIPLE_SUBSYS T

in &MULTIPLE_FORCE_EVALS .. &END

 Kind regards

Marcella

[daniel Storm]

Hi Marcella,

Thanks for your response.

I tried including MULTIPLE_SUBSYS T in the MULTIPLE_FORCE_EVALS section and unfortunately that gives me the same error as before.

I have also tried varying the size of my QM region. This system has 4 cations and 4 cations in it with an overall charge of 0. I’ve tried having my QM region as;

-              -   1 cation (Charge 1)

-              -   1 cation + 1 anion (Charge 0)

These unfortunately produce similar errors to above.

Kind regards,
Daniel

[Marcella Iannuzzi]

Shouldn't you pass to each FORCE_EVAL its own coordinates file?

See the test cp2k/tests/FE/regtest-1/Solv_alch_chng_simpl.inp

In your input it seems that each FORCE_EVAL gets the same coordinate file

Kind regards

Marcella

[daniel Storm]

Hi,

Sorry for the late response. I’ve done some testing and implemented what I could from the input file you suggested, and I got the calculation to run however I’m now getting different issues. The problem is during the optimisation. The molecule doesn’t stay together properly, Hydrogens fly off in random directions and other atoms can overbind into a sort of ball. I’ll make a list of all the things I’ve tried below.

What I’ve tried:

-          - Periodic unit cell (currently in .inp)

-          - Non-orthorhombic parameters (currently in .inp)

-          - Non-periodic unit cell

-          - Orthorhombic parameters (40 40 40, 90.0 90.0 90.0)

-          - Reduced the size of the system to just 1 cation (QM) and 1 anion (xTB)

All of these give me the same errors. My other thought was that it could be an error in the basis set and ECP section or it’s in the wrong place? But I’m not so sure. Looking into the error messages of ir_tBu2PONOP_meh_oniom_27-r-3.out it crashes on FORCE_EVAL 4 which is where xTB is used on the QM region.

I’ve attached all relevant files again and any help would be much appreciated. One thing to note is that I’ve found a different approach to use for this project which seems to be working. I’m more looking into ONIOM now for interest, so there’s no problem if it can’t get sorted.

Thanks,

Daniel