CIF file types - differences?

Hi!

I’m using the materials project for my research project at uni, and I was wondering if there’s an easy explanation of the differences in the CIF files that you can download? Like, what the conventional standard CIF does to the structure?

Hope that makes sense, just getting myself a bit confused on which I should use as I’m not sure what the differences are

Thanks!

I may have accidentally posted this twice - sorry!

Hi @Park,

The most important thing is that all the CIFs are either exactly or effectively equivalent, they are just different “settings” of the same crystal structure. So shouldn’t matter too much which you use.

These are:

  • Computed, the structure direct from our calculations, always in the “P1” setting (i.e. no symmetry is detected)
  • Conventional standard, this is the same as “computed” except now the symmetry is detected and it’s put into a conventional setting – this will be more familiar to most scientists, since textbooks and papers etc. will usually report a crystal in its conventional setting
  • Primitive cell, this is the same as “computed” except now it’s reported in its primitive setting (i.e. with the minimum number of atoms required to define the periodic crystal). It’s often useful to use these primitive cells for further calculations, since they contain fewer atoms, but they can be more difficult to visualize – for example, a crystal with cubic symmetry might not have a primitive cell whose angles are 90º. Introductions to crystallography can go into more detail on this.
  • Symmetrized is the almost the same as “computed”, except atomic positions have been moved to their symmetrically equivalent positions if they’re almost but not quite there – this is important e.g. if due to numerical noise an atom may be close to but not exactly at (0, 0, 0) even if we know in reality it’s likely to be exactly at (0, 0, 0) in that particular setting.

Best,

Matt

1 Like

Ah that’s great, thanks so much! :grinning:

I really appreciate you taking the time to explain it for me @mkhorton !