[GSAS-II] GSAS-II and Origin 1 (where a choice)

[Toby, Brian H.]

Just to set the record straight, GSAS-II (and GSAS) will always use Origin 2 settings for space groups where there is a choice. Origin 1 does not put the center of symmetry at the origin, which means that double the amount of structural computations must be done and reflection phases are complex quantities — both completely unnecessary. It will be clear when Origin 1 coordinates are fed into GSAS-II because the computed site multiplicities and the density will be wrong. GSAS-II does have an option in the transform menu to convert coordinates from Origin 1 to Origin 2 for the small number of centrosymmetric space groups where the highest symmetry point in the unit cell does not have the center of symmetry. However, the software cannot recognize that the wrong structure is being generated; a human needs to look at one of: site multiplicities/density/structural plot/… and decide something is wrong.

Wen Wen is correct that in other programs that have both sets of symmetry operators as options, one gets the same structure regardless of origin, but even then human intervention may be needed to get the origin selection right. Even CIF did not originally handle this very well.

Brian


On Aug 26, 2019, at 4:48 AM, wenwen--- via GSAS-II <gsas-ii at aps.anl.gov<mailto:gsas-ii at aps.anl.gov>> wrote:

From my understanding, the origin should not be a problem.
For standard Cu2O with space group of pn-3m, the atomic sites for both Cu and O should be well set already. However, the euivalent sites for both oxygen/or Cu should be carefully considered if you shift the orign of the molecules.
origin 1 Cu 4b（0.25 0.25 0.25） and O 2a (0 0 0)

<Catch.jpg>
origin 2 Cu 4c (0.5 0.5 0.5) and O 2a （0.25 0.25 0.25）

<CatchB8E0.jpg>


the previous 4b sites changed to 4c, instead.

Correct me if I am wrong.

Best regards,
Wen Wen


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.aps.anl.gov/pipermail/gsas-ii/attachments/20190826/97c48a93/attachment-0001.html>