Re: (Rietveld Calc Data) vs (ICDD Ref Data lacking trace peaks)

Armel Le Bail ( armel@fluo.univ-lemans.fr )
Fri, 27 Jun 1997 09:07:08 +0200

Lachlan wrote:
>I have been giving some commercial Powder XRD training seminars -
>including search-match on PDF-1 database using commercial software
>the laboratories have purchased (in this case Philips PC-Identify).
>I am curious about the following and am interested in other know
>of similar phases and/or want to trade some phase ID pitfall stories.

Lachlan, you know that this is one of my preferred orientation :-).

2/3 of my stories are pitfall stories. No place here to give details. The
equation is simple :

65000 cards in PDF-2,
200000 phases characterized by their crystal structure, or more.

Maybe I am not a lucky man because I had to identify several of the
135000 compounds lacking in the database but this is not a question
of chance, this is a consequence of the antediluvian ICDD politics.

You know that we have changed our government here in France. A
new Minister said that he wanted to "take the fat from the mammoth" ,
speaking of the National Education. ICDD is a kind of fat mammoth
with 200 rules giving reasons why the 165000 compounds are not there.
Adding the peoples at the Board of Directors + the Staff + the 100
or so "Active ICDD Members" gives an idea of the freezing degree of
this organization adding 2000 cards in the bank each year when the
Cambridge Structural Databank adds 10000 entries (organic +
organometallic), the ICSD adds 1000 (?) or so entries and CRYSMET
adds (???) metallic compounds.

One of the 200 rules is presumably "never trust a crystal structure, don't
add a calculated pattern in the bank, preferably"

My conclusion is : instead of adding more and more (in)Active ICDD
Members, please add calculated patterns.

Lacking trace peaks is a negligible detail if compared to lacking 2/3 of
the posiible data !

(OK, I said that before:-)

Armel