Optimizing with TIP3P a water cluster model

[Vittorio Bariosco]

Dear CP2K users,

I have been through different threads on this forum to understand how to optimize a cluster of 600 hundred atoms (see 0_tip3p_geom.xyz) with TIP3P model.

1)I firstly tried this input taken from the FIST repository on GitHub here

(see first_tip3p.inp) but I was always getting this error after the Optimization starts:

"WARNING| Particles:     600    598 at distance [au]:     1.82837959 less than:      2.00430963; increase EMAX_SPLINE.

GEOMETRY wrong or EMAX_SPLINE too small!"

2)The I moved to the other input reported in the same repository at this link, where more water molecules were included in the initial geometry. In this case (see second_tip3p.inp),  the optimization starts but at a certain point it crashes again with a similar error: 

"WARNING| Particles:     600    598 at distance [au]:     0.89796367 less than:      0.95400000; increase EMAX_SPLINE."

GEOMETRY wrong or EMAX_SPLINE too small! 

At this point I am a little confused on how to proceed. I remembered that tip3p is a rigid FF and I started to look for constraining options for intramolecular distances and angles. I found in the same input of the above the section &CONSTRAINT &G3X3.

3) Thus, I defined the new input using this constrain that could probably explains the reason why CP2K was crushing before in this way (see third_tip3p.inp). However, also this test was unsuccessful with the following error:

"    |  Error in constraints setup!

    |  A constraint has been defined for a molecule type

    |   but the atoms specified in the constraint and the atoms defined for

    |   the molecule DO NOT match!

    |  This could be very probable due to a wrong connectivity, or an error

    |   in the constraint specification in the input file.

    |   Please check it carefully!"

Moreover, I also found in another thread here, that the specified constrained keyword works only for MD runs and not for GEO_OPT or CELL_OPT. 

At the moment I don't have any clue in how to proceed to obtain a optimized geometry with tip3p model. For this reason I would like to ask if anybody have ever optimized with tip3p model, or can help me for setting the proper input.

Hope to hear from you.

Kindly,

Vittorio

[Krack Matthias]

Dear Vittorio

 

a CP2K input file for 32 TIP3P water molecules is attached.

I plan to provide template CP2K inputs for the common 3- and 4-site rigid and flexible water models in the CP2K data folder soon. I will announce that here when done.

 

HTH

 

Matthias

 

--
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 view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/e32370de-519b-4fd3-9176-438ddf027704n%40googlegroups.com.

[Vittorio Bariosco]

Dear Matthias,

Thank you very much for your answer. I will surely try it.

Best,

Vittorio

[Vittorio Bariosco]

Dear Matthias, 

I tried with the same input substituting to the MD run a GEO_OPT. However,  it gives back again the error at 8th optimization step:

"GEOMETRY wrong or EMAX_SPLINE too small! "

Maybe is not possible at all to use tip3p for optimizations?

Kindly,

Vittorio

[Krack Matthias]

Dear Vittorio

 

The G3X3 constraint, required for rigid 3-site water models, is not implemented in CP2K for the run types GEO_OPT and CELL_OPT.

 

Best

 

Matthias

 

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/d50f18c7-8abc-4d98-9129-c87c25ee4870n%40googlegroups.com.