I'm working on Iron phases, and performing Rietveld analysis routinely, as
part of catalyst kinetics studies which make up the major portion of my
dissertation. I nearly went INSANE, initially, trying to ID magnetite using
Search-Match, trying to cross-reference with the refinements. I've using
combinations of ICDD, Neutron and XRD data, Search-match routines, Fink and
Hanawalt ID methods, reflection data, TEM data, and simulated patterns from
single crystal and standard patterns that I ran with materials we've bought,
materials we've 'borrowed' and materials we've made, amongst other things.
How much work and effort I put into a pattern is governed by what it's being
used for.
The first thing that I found was that I could match some peaks, but not all,
using ICDD PDF-1 and -2 data. I also got many mismatches. Which ones I
could match depended on which cards, including discarded data, I looked at.
Then there are the accursed trace peaks. Alas, I know them well (my
apologies to Shakespeare!).
I'm very familiar with the magnetite peak at 1.929 =C5 - I spent a couple=
of
weeks convincing myself it was really a magnetite peak, when I first began
work on this project. My sample was 'home-made' and pure as proved by other
techniques, and I wouldn't accept the possibility of an impurity, in spite
of what I was being told by more experienced heads (at the time) than mine.
I started out as a single crystal person and suffered with powder patterns,
until I realized that the reference data base was neither accurate nor=
complete.
I analyze all kinds of material besides my project samples, usually for
others in my research group, and have found, the hard way, also, that I
can't rely on search-match, or on ICDD - I routinely use Rietveld as a tool
for phase ID and often use materials that I consider, after indexing by
hand, and running to the literature, to be isostructural. Being able to
quantify AND identify at the same time is lovely. I also catch ID mistakes
very quickly, and solve questions about trace peaks, with the least amount
of pain and hair loss (I pull it out a strand at a time when stressed ...
sounds like you do, too). I've also found that I'm seeing the same degree
of detail (e.g., trace peaks, minor phase relative intensities) in my data
whether I use fast-scan, for phase ID, or collect Rietveld-quality data.
I never trust anyone who gives me a sample, anymore - learned that one the
hard way, also, with CA, Al, Mg, and Pb phases, for instance. I just this
week ran a set of 4 'identical' (3 synthesized) zeolite samples which turned
out to be 1) MAP zeolite, 2)cristobalite (don't ask me how THAT happened),
3) a silica-alumina-calcium oxide mixture, and an as-yet incompletely ID'd
zeolite mixture. They're supposed to be MAP. The first was the standard.
The results using PDF-1 search-match were - interesting, to put it mildly.
In addition, the particle size distribution of the powders affects the peak
intensities/ratios I've been detecting, using the same collection
conditions. Preferred orientation in samples can cause serious mismatches,
which are avoided in rietveld refinement. I'm also finding that some phase
mixtures can cause 'unusual' intensity enhancements and extinctions which
can muck up the works with respect to routine ID.
I don't know of a 'cheap' route to PDF-2 data from PDF-1 - we've looked into
it, also. One can use just PDF-2 data for phase ID of complete unknowns,
although I tend not to, but I would not want to rely on 'just' PDF-1 search
match and I tend to advise people not to, unless they're experienced
crystallographers. At least, as you say, the analysis and synthesis
conditions are available with PDF-2 and one can make informed decisions. I
know individuals who've erred grieviously using PDF-1 data.
I'd love to see the ICDD databases updated/corrected for errors and
reconciled with the literature to include several forms of data for a given
compound, as you say. Lower costs would be nice also ... =20
I'm not familiar with'GRANT-IN-AID' patterns - I assume these are known,
reference samples? For knowns, I do sometimes use rietveld simulations to
determine scan conditions.
Forgive me, I've been helping to write grant proposals, and am still in
'justification' mode. I'll stop babbling now. I'm more than willing,
obviously, to continue a discussion of these things.
Cheers,
Linda
Linda D. Mansker EMAIL: mansker@unm.edu
Chemical and Nuclear Engineering (505) 277-5841
University of New Mexico FAX: (505) 277-5433
% % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % %
"Out of clutter, find simplicity. From discord, find harmony.
In the middle of difficulty, lies opportunity."
Albert Einstein
% % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % %