From:   GORDON::COCKCROFT    "Jeremy Karl Cockcroft (Tel. 0171-631-6800)" 15-JUN-2000 17:47:50.69
To:     SMTP%"S.F.Parker@rl.ac.uk"
CC:     COCKCROFT
Subj:   RE: Indexing program

Dear Stewart,
   The indexing program in the course makes no assumption about symmetry
and effectively assumes that everying is triclinic and looks for the
symmetry after finding a possible unit cell. The best check is to take
the suggested unit cells, and feed the output as input to an hkl generating
program. Compare the generated reflection list to the data - visually
if possible - the human eye is better at pattern patching than most computer
programs. To answer your first question, the indexing program may say
something is monoclinic when it is orthorhombic, especially if the
orthorhombic cell is centred (though it will attempt to check for this).
Some indexing programs make assumptions about the crystal system, but are
then slow when the wrong assumption is made.
   All indexing programs, irrespective of method, work best when the longest
d spacings are given since these generate the shortest distances in reciprocal
space. Having said this, some will work well when these are not available
and others won't (e.g. the Taupin type method works very badly if the
long d spacings are missing, whereas zone indexing still works moderately
well). One of the problems with indexing programs (and I can't remember
whether I mentioned it in the course material - I meant to do so in
case I didn't) is that of zone dominant planes due to two large unit cell
lengths and one small one. This can result in the largest d spacings being
all of the type hk0 in which case nonsence values for c may be obtained.
   It is often worth trying several indexing programs if possible. The other
point is that even if 20 lines are indexed, it is still possible to have
the wrong unit cell, especially if there is any pseudo symmetry in the
system.
   I hope this has answered your questions. If not, get back to me again.
          Cheers, Jeremy Karl.
P.S. I accept your point about inputting d spacing values. It is quite
hard work to get these sort of things tested though. You would be amazed
at the number of man hours that goes into putting on this type of course.
I hope to get the next two sections of the course material out shortly.

Back