Dear Mr. Le Bail
again I have to point Your attention to the fact, that the PDF products of the
ICDD are compiled for phase identification in the field of condensed matter. It
was not the basic idea to collect all data of crystalline phases. And as a
further point, one should keep in mind, the PDF database covers a wide range of
interests (metallurgists, industrial and applied mineralogist, geologist,
forensic sciences, etc.), not only oganic and organometallic chemists.
1. Of course we all dream of the 'Eierlegende Woll-Milch-Sau' (roughly transated:
egg, milk, and wool giving hog), however, the problem is not the ICDD, the
problems are the economic interests of ICSD and CSD. So if You hestitate about
ICDD policy and practice, why dont You hesitate about the other two database
vendors.
2. With respect to inorganic chemists, the problem is solved since release 1998.
And after the incorporation of the CSD data, the PDF will cover Your
requirements.
text.ori > You are a representative ICDD member which
text.ori > was happy with that PDF-2 database containing 59800 "active
text.ori > patterns" in the 1994-95 release, up to set 44. You apparently
text.ori > do not care about the notion of database completeness that is
text.ori > a constant dream of analysts who observed >200000 crystal
text.ori > structures determined at the same time. Thus, ICDD
text.ori > represented less than 1/3 of what it should have been if a
text.ori > correct job had been made.
3. This is the wrong way of thinking. If I want to change a system, I can change
it only from inside, not from outside.
text.ori > I am not a member and was and continue to be unhappy with
text.ori > PDF-2 (though, ICDD proposed to me several times to become
text.ori > so, and the first time they propose was the first time I critisize
text.ori > PDF-2, in a paper submitted to Powder Diffraction -1994 - which
text.ori > had difficulties to be publish, hence my permanent angry - I may
text.ori > resort the referee comments, just for kidding more ;-). I have
text.ori > some doubts in the efficiency of an organization obtaining high
text.ori > percentage agreement from its members by fast hands up voted
text.ori > projects, as I could see at the Denver X-ray Conference, 1998.
4. It is up to everyones mind, to establish his own specific database. It is the
usual way to speed up the phase identification process in any routine industrial
application. But I think it is a waste of time to establish a second PDF
database.
text.ori > I guess too that I will be unhappy with PDF-3 and try to
text.ori > produce an alternative for free :
text.ori > http://sdpd.univ-lemans.fr/powbase/
5. Industrial scientists are usually not interested in structural data, their
main interests are quality control and qualitative raw materials
characterization. Think of cement industries.
text.ori > I do not understand how industrials can be satisfied if academic
text.ori > researchers are not. But I can see how the "nonprofit" ICDD
text.ori > could consider not necessary to give satisfaction to academy if
text.ori > industrials are the main clients (how can you expect that an
text.ori > analyst working for an industrial could complain ? He would
text.ori > be fired...).
6. DVDs for an economic rate are only available since 1 year.
text.ori > once. DVD technology allows the storage of the whole stuff
text.ori > on one disc.
7. What I ment, there are a lot of problems which can be solved more easily or
only with single crystal divices. Think of all that unstable organic and
organometallic crystals, stable only for several hours or under Ar.
8. I would not be happy with a unified database, with all the specific
informations of all three databases. A market without competition, will reduce
the quality, see Microsoft.
9. Of course if You have a database of 60000 entries, You will always will find
problems like You give below. However, it is on behalf of us, the scientists, to
report these problems to the editorial staff of the ICDD. And dont forget all the
mistakes entered in the ICSD and CSD. Additionally just remember a superb
datatypist will have an error rate of < 1%. And now think of the 400 MB
characters and numbers of the PDF2.DAT file. If You hesitate about the quality of
literature data, the only thing I can say, they reflect the quality of the work
of our colleagues, not the quality of the ICDD staff.
text.ori > Clearly, more than 70% of the 59800 patterns proposed in
text.ori > 1994-95 were not submitted to that quality control process.
text.ori > I can show you an example taken in PDF-2 each day up to
text.ori > my retirement. Let us just start with the files concerning
text.ori > [Pd(NH3)4]Cr2O7. Lot of info about it :
text.ori > 39-1422 triclinic cell, huge preferred orientation
text.ori > 40-1486 was exactly the same as 39-1422 up to the
text.ori > structure publication from powder data. Then,
text.ori > that file was modified in a curious way.
text.ori > The correct monoclinic cell was added, removing
text.ori > the false P1. Accordingly, the new indices hkl
text.ori > are given. But the old Iobs and dobs are kept,
text.ori > and used to recalculate an F30 = 6 which is
text.ori > quite bad if compared to the F value given in the
text.ori > publication. But ICDD did not wanted of the
text.ori > dcalc and "Iobs" produced by the Rietveld method.
text.ori > They absolutely preferred those badly estimated dobs
text.ori > and Iobs produced by peak fitting or second
text.ori > derivative method. The paper authors (including
text.ori > me of course) said that they would not deliver such
text.ori > values because of the compound complexity and
text.ori > high overlapping degree which make them inaccurate...
text.ori > 85-1942 : fortunately, that card from ICSD characterizes now
text.ori > well that compound.
Finally as I pointed out for several time: The PDF is an identification tool in
the field of condensed matter on the basis of observed powder patterns. It would
be useful to have link-like information to the ICSD and CSD data, however, as
these products cover different data they should be compiled by different
specialists. The preferrence for observed data is based on the reasonable
assumption that a compound measured under standardized conditions, and evaluated
with similar algorithms, should provide the user with results compareable to the
data compiled in the PDF2 file.
With greetings from Vienna
Christian Lengauer
#
#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~#
# #
# Christian L. Lengauer #
#
