Re. calculation of IR spectra
As indicated in a previous message, the crystal structure only determines part
of the problem of calculating vibrational spectra - the other part is a
knowledge of the interatomic forces, and this is surely the hardest part.
There is quite a lot of expertise around on the calculation of vibrational
spectra. In our group we are always interested in the vibrational spectra of
ionic materials (e.g. LiNaSO4 is the material we did some calculations on
yesterday) and particularly in silicates. For the latter, we have some very
good models that can be applied (so it appears after some detailed testing) to
a wide range of silicates (examples, silica polymorphs, garnets, feldspars). We
use Julian Gale's GULP program, which gives mode frequencies and IR
intensities. This is a very nice program, BUT, the whole approach is not a
simple black box that can be tinkered with to give definitive answers without
thought and experience. That experience is required to get the most out of
Rietveld analysis applies to modelling also!
I have some more information and leads in our www home page
(http://www.esc.cam.ac.uk/astaff/dove/), but this may provide a reason to
update it!!!
Martin