#188 load molecule

None
closed
nobody
None
5
2013-03-26
2011-03-23
Simon Westrip
No

Attached CIF when opened in Jmol 12.1.40 prerelease version does not create a connected molecule for one of the moities - i.e.
the CIF bonds include C9-C9(−x+2, −y+2, −z+1) but this bond (and the symmetry-related site C9) does not seem to be rendered,
though the 'connected' symmetry-related sites appear (thus forming two partial molecules - missing the connecting atom).
Is there any way round this?

Discussion

  • Simon Westrip
    Simon Westrip
    2011-03-23

     
    Attachments
  • Bob Hanson
    Bob Hanson
    2011-03-24

    this is a bug, not a feature request. Fixed for Jmol 12.1.41 and 12.0.37

     
  • Bob Hanson
    Bob Hanson
    2011-03-24

    make that just 12.1.41 -- this molecular load of CIF files is not part of 12.0.

     
  • Simon Westrip
    Simon Westrip
    2011-03-29

     
    Attachments
  • Simon Westrip
    Simon Westrip
    2011-03-29

    Thank you for foxing this.

    I have a similar problem with the attached CIF (wm2328d.cif) where a naphthalene ring is 'divided' by symmetry.
    Could the 'molecular view' be enhanced to handle such cases?

    Cheers

    Simon

     
  • Bob Hanson
    Bob Hanson
    2011-03-31

    Not in this case. It's a polymeric molecular solid, like diamond, not a set of molecules. So no matter how you do it, you would have to clip it somewhere. In this case, Jmol simply delivers the atoms as they are listed in the CIF file.

    I'm having trouble installing Mercury, or I would check how they handle this.

     
  • Bob Hanson
    Bob Hanson
    2011-03-31

    I do see that Mercury is able to decipher that and keep the naphthalenes together. Beats me how they do that. It's certainly a pretty good trick.

     
  • Simon Westrip
    Simon Westrip
    2011-04-02

    Thanks again for looking at this. I dont know how Mercury approaches this, but I had a look at PLATON.
    The script below seems to emulate the PLATON approach. It doesnt produce an ideal view, but then
    again its perhaps very difficult to say what an 'ideal' view is in such cases.

    Anyway, hope this is of interest.

    load "" {444 666 1}
    hide none;
    select not symop=1555 and connected(symop=1555);
    var ss={selected};
    select symop=1555;
    for (var s=0;s<ss.length;s++)
    {
    var x= @{ss}[s].symmetry;
    x="symop="+x;
    x = x.replace(",",",symop=");
    echo @x;
    select {selected}, @x;
    }
    hide not selected;
    select none;
    unitcell off;
    axes off;

     
  • Bob Hanson
    Bob Hanson
    2011-04-02

    That's still not it. I see Mercury's load now. Looks like they are ensuring that all organic parts are continuous and allowing the model to break along metal bonds.

    I guess there is comfort in seeing that Mercury also gets it wrong in a different way -- not indicating the alternative locations on some atoms, and showing two Mg for each S rather than 1.

     
  • Simon Westrip
    Simon Westrip
    2011-04-11

    I've had a go at scripting a 'Mercury' view, but without much success. However, in some respects the Mercury view isnt ideal for these polymeric organometallics. So I've scripted a 'coordination' view that displays the ligands in their entirety around the metal sites and cuts off the view where these ligands bridge to symmetry-related metals. Such a view at least avoids 'broken' ligands.

    Some examples are at http://journals.iucr.org/testing/chemistry/ (follow the 'Supplementary page' links - 2nd example shows a coordination polymer).

    Again, hope this is of interest.

     
  • Bob Hanson
    Bob Hanson
    2013-03-26

    • status: open --> closed
    • milestone: -->