[TCOD] [SPAM] Re: What structures do we accept to TCOD?

[Björkman Torbjörn]

Dear all,

It seems that I forgot to add the list to recipients... Armel, do read on, I added some things.

I would say no, for two reasons. The first is practical: It would require the possibility to somehow automatically determine a suitable experimental reference structure. I don't think that this is possible, since the criteria go way beyond "purely crystallographic" considerations (Fobs et cetera). You'd need to know that the sample is actually pure enough (in DFT we always have perfect samples...), precisely how were samples prepared, what the temperature effects are like and so on...

The second is from basic principles. I think that would be going into data analysis, whereas the point of TCOD is just storage of data, with quality control done within DFT itself (to the extent that this is possible). Requiring a comparison to experimental data to me makes about as much sense as requiring supporting DFT calculations for inclusion of experimentally determined structures in COD. All the arguments against such a procedure apply equally well in both directions.

There is also the problem that, in terms of Saulius checkpoint list, we also have the possibility
b''') Neither experiments or theory have any problems, but you are comparing apples and bananas due to too little information being available on either side. 
For example, a DFT calculation (which assumes a temperature of 0K) would tell you that the experimentally observed bcc titanium is nonsense, whereas it is in fact a high temperature phase that is stabilized by phonon-phonon interactions. If temperature data would be missing (or, worse, being set to some default ~300K), this would generate confusion. The only time that this is straightforward is in cases where a lot is known and everyone knows the correct answer anyway, but then these are the cases we are trying to venture beyond...

It would of course be very nice to at some point have a "gold standard" database, containing only structures that are exceedingly well known and understood both from the points of theory and experiment, but I don't think that either COD or TCOD should be that database.


Regards,
Torbjörn

---
Torbjörn Björkman, PhD
COMP, Aalto University School of Science
Espoo, Finland


________________________________________
Från: tcod-bounces at lists.crystallography.net [tcod-bounces at lists.crystallography.net] för Armel le Bail [Armel.le_Bail at univ-lemans.fr]
Skickat: den 30 juli 2014 01:02
Till: tcod at lists.crystallography.net
Ämne: Re: [TCOD] [SPAM] Re:  What structures do we accept to TCOD?

Hi,

> My point is: the task of comparing experiment and theory (i.e.
> comparing COD and TCOD) is a research task in its own right.

To my knowledge, theoricians are aware about the low quality of the
cell parameters derived from their DFT optimizations to the point that
they prefer to fix them to the experimental values when available.
There is here at least a point of convergence between COD and TCOD -
and a serious problem with the current theoretical approach ;-).

Best,

Armel



_______________________________________________
Tcod mailing list
Tcod at lists.crystallography.net
http://lists.crystallography.net/cgi-bin/mailman/listinfo/tcod