In our lab
we usually have to analyze geological samples for fluorine. The normal
- low F amounts in Fe-rich
samples (e.g. biotite, with 20% FeO and 0.5% F).
- high F contents in beam-sensitive
Fe-rich samples (e.g. triplite, a phosphate with 5% F and 19% FeO; see
Fialin & Chopin 2006, American Mineralogist 91: 503-510).
contents in Fe-poor samples (e.g. 3.5% F and 0.5% FeO, as in apatite) is less of
a problem, although still rather common since this mineral is widespread.
crystal gives a very low count rate, which is a problem for beam-sensitive
samples which need rather short count times (around 10 s, and it would be
better to shorten this time); in addition, these materials cannot withstand
very high currents without F migration, which results either in an increase or
decrease of the count rate (probably depending on the crystal orientation with
respect to the beam). Moreover, higher-order P and Ca lines interfere with F
(they can be filtered with PHA, but that´s an additional complication and leads
to even lower counts). The LDE1 crystal, whose count-rate is with several
orders of magnitude higher, has severe F-Fe overlap and troublesome background
positions (reasonably close to the Fe+F peak) in samples that have Mn and Al.
wondering if you would mind to share with the list how you cope with this problem
in your lab. Has anybody adopted the approach of Witter and Kuehner (2004,
American Mineralogist 89: 57-63)? If so, are you satisfied with the results?
advance for your input,
Nacional de Córdoba