Dear Peter and all,
I obviously totally agree with you concerning the
absolute necessity of good petrographic
observation.
It can happen, though, that even with good
petrograpic observation you sometimes don't know
what you are dealing with (in particular since
convergent light microscopy is mostly not taught
anymore or at least most people have difficulties
using it in a routine way).
The first time I found large crystals of
wagnerite (Mg,Fe)2(PO4)F in one of my rocks, I
was not able to identify it optically - and then
webmineral.com was useful for interpreting the
EPMA data. Incidentally, this experience, in
particular analysing the F (vs. OH) in wagnerite
made me realise the problems with standards for
fluorine and the dependence of the analysis on
the crystal orientation that you mention...
All the best,
Pavel
At 14:31 +0000 15/01/16, Treloar, Peter J wrote:
>Dear All
>
>I read Pavel's comments with interest and felt
>that I needed to add something here. There might
>be a bit of a rant here - but please read on and
>enjoy it.
>
>I have just taken early retirement after a life
>time working as a mineralogist and petrologist.
>When I did my first degree in Bristol we were
>taught how to use the optical microscope as the
>primary means of identifying minerals. I first
>used the EMPA as a PhD student and then spent
>four years as a post-doc in Jim Long's lab in
>Cambridge running the then state-of-the-art WDS
>and EDS microprobes. I have followed with real
>fascination how these various have evolved and
>have been involved in buying three SEM based
>EPMA systems that deliver both WDS and EDS data.
>We have widely published precise, accurate data
>from these instruments and I will argue to the
>end that EDS is unfairly criticised by those who
>do not understand the counting statistics.
>
>WDS is no different now. Count rates are still
>constrained by flow counters but it is now all
>computer driven but you still need to know what
>you are doing. Peak searches, P/B ratios etc,
>proper standards.
>
>EDS is more interesting. Peter Statham (who cut
>his teeth in Cambridge in developing the first
>EDS system with Jim Long and is no at Oxford
>Instruments) said to me a few years ago that "we
>now finally understand the physics of the EDS
>spectrum and how to resolve the overlaps".
>People are still "sniffy" about EDS data - but
>it is great. With the new Pentier cooled EDS
>detectors (which do not need liquid Nitrogen)
>count rates are faster, P/B rates are better,
>precision is better and mapping is faster. The
>growth of non quantitative EDS mapping of
>mineral assemblages reflects this. For those who
>wish to ask -about Pentier cooled detectors I
>did test this when we bought our latest system
>and they do work. Peak shift and peak broadening
>is no longer an issue with these systems.
>
>EPMA today begs new questions. How best can we
>(if we can at all) generate accurate (precise is
>potentially not possible) analytical data using
>a variable vacuum SEM with no carbon coating of
>the target? Our data look good - if not
>encouraging.
>
>What effect does the sample orientation have on
>analytical totals? Rephrase that into a
>discussion of electron channeling and then think
>of an orthorhombic or tetragonal mineral.
>
>BUT and this is a BIG BUT: I know that there are
>minerals that you can not easily identify using
>optical techniques. But the bottom lone before
>any scientist puts a sample into an EPMA they
>need to know what they are looking at. This is
>basic science. Why spend grant money identifying
>quartz or calcite (or even differentiating
>between pyroxene and amphibole) when you should
>be able to do it on the optical microscope.
>There is a need to do optical microscopy not
>just to identify mineral assemblages, but do the
>whole paragenesis thing.
>
>As Pavel says - MINDAT is the best of the
>various chemical packages that you can use. BUT
>- please do the optical mineralogy.
>
>After all:
>
>You do need to use chemical stoichiometry to
>differentiate between so why not start with the
>optics.
>
>My final message is to anybody who thinks that
>they can cut corners: it is not just us old
>guys, but the best petrologists working today do
>optics first, second and third and then, and
>only then, EPMA.
>
>Regards
>
>Peter Treloar
--
Pavel PITRA
Géosciences Rennes email: [log in to unmask]
Université Rennes 1 tel: (++33) 2.23.23.65.06
Campus de Beaulieu - Bat. 15 fax: (++33) 2.23.23.60.77
F - 35 042 RENNES CEDEX
FRANCE http://www.geosciences.univ-rennes1.fr/
http://www.geosciences.univ-rennes1.fr/spip.php?article67
|