Date: Wed, 1 Jun 94 12:06:55 EST
Subject: Chemical shift in x-ray spectra
From: Nick.Ware@anu.edu.au

Thanks to those subscribers who have discussed the effects of chemical
bonding on x-ray spectra. I feel that it would be refreshing to do some
real chemistry with the electronprobe instead of churning out endless
quantitative analyses!

The modern microprobe uses optical encoders to position the spectrometers
and can reproduce digitised spectra with an ease unimagined by the
researchers who did most of the EMPA chemical spectroscopy back in the
1960s. There must be some inorganic or organo-metallic chemists out there
who could use decent x-ray spectra to advantage.

Unfortunately, previous work has been directed to solving problems better
tackled using boring old conventional analyses. For example,the Fe++/Fe+++
ratio in oxides may be calculated by looking at the Fe L-spectrum where the
L-beta peaks can be more intense than the L-alpha in certain cases.
However a good quantitative analysis can give a better Fe++/Fe+++ ratio
either by using stoichiometry constraints or by analysing oxygen directly.
Again,the Al K-alpha / K-beta ratio in certain silicates may be measured
and the ratio of aluminium in the 4-fold and 6-fold sites calculated.
However, just as good a value may be obtained from a quantitive analysis
from which a "structural formula" can be calculated.

To use the technique to advantage you need a researcher primarily
interested in what the electrons are doing in the compound she/he has made.
Such people ought to exist in chemistry departments but they seem to be
few and far between.

Nick Ware

N.G.Ware Phone (06) 249 3429
Research School of Earth Sciences A.H. (06) 295 2219
Australian National University Fax (06) 249 5989
Canberra A.C.T. 0200