GEOMETRY wrong or EMAX_SPLINE too small!

[Henrique Junior]

Doing my first steps with cp2k I've bumped in this error: GEOMETRY wrong or EMAX_SPLINE too small! (files attached).
My intention is to use cp2k to calculate the magnetic coupling in solid state between two coppers (by evaluating different possible distances of Cu-Cu interactions).
(I'm not trying broken symmetry yet, this is just my first attempt).

Thanks in advance

[Rolf David]

Hi,

Too atoms are too close ?

WARNING| Particles:      96     25 at distance [au]:     0.00075400 less than:      0.01889726; increase EMAX_SPLINE.

When i encounter this error, i'll change the geometry of the involved atoms and it goes smoothly after.

[Henrique C. S. Junior]

Thanks, Rolf,

In this case, the structure was determined using X-Ray diffraction and hydrogens are optimized via DFT, so, the structure is probably correct. Is there a way to fix this problem wothout changing my structure?

By the way, I've used Avogadro2 to generate my input (so I don't have to write it from scratch).

----------

Henrique C. S. Junior

Químico Industrial - UFRRJ
Mestrando em Química Inorgânica - UFRRJ

Centro de Processamento de Dados - PMP

---

Date: Sat, 18 Jul 2015 01:46:08 -0700
From: rolf.d...@gmail.com
To: cp...@googlegroups.com
Subject: [CP2K:6794] Re: GEOMETRY wrong or EMAX_SPLINE too small!

--
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.

[Matt W]

Hi, 

maybe you've got your lattice parameters incorrect then? Two atoms are on top of each other with.

Matt

[Henrique Junior]

Thank you for all your patience answering my questions, friends, but here I go again:

I'm testing with another structure (attached, determined by X-Ray diffraction too). Looking at the manual I've tried this (with arbitrary values like 9.0, 99.0, 999.0 etc for EMAX_SPLINE just for a try):

&FORCE_EVAL

 &MM

   &FORCEFIELD

    &SPLINE

     EMAX_SPLINE 99.0

    &END SPLINE

   &END FORCEFIELD

 &END MM

...

The error persists:

WARNING| Particles: 30 17 at distance [au]: 0.00113006 less than: 0.01889726; increase EMAX_SPLINE.

******************************************************************

*** ERROR in build_neighbor_lists (MODULE fist_neighbor_lists) ***

******************************************************************

*** GEOMETRY wrong or EMAX_SPLINE too small! ***

*** Program stopped at line number 654 of MODULE fist_neighbor_lists ***

The distance that CP2K is complaining about (0.00113006 au) is equal to 5.98001996e-04 A (very, very small) but as far as I can see, my coordinates havent changed in my input (compared to my .cif file) and everything is OK.

I have deleted some units from my unitary cell because I'm using FPBU to test specific Cu—Cu distances.

--

[Matt W]

Hi,

I think the lattice parameters are off - you give a set on nonorthogonal lattice vectors and then you specify the angles - I have no idea what this does.

Try with just the lattice vectors or giving ABC and alpha_beta_gamma

Matt

[Henrique Junior]

Dear Matt, looks like you're right. After changing my lattice parameters to a arbitrary value like 10.0 10.0 10.0 the error was fixed. I'll have to take a closer look at my parameters.
Thank you.