Date: Wed, 28 Sep 94 13:37:51 PDT
Subject: Re: Thanks and More on Quant EDS
From: tonygr@EAGLE.MIT.EDU (Anthony Garratt-Reed)

You don't eliminate absorption effects by using the K and L lines of
Ni metal films - its is just that the self-absorption of Ni-L in Ni is much less
than, say Ni-L in Co or O. For example, some mass absorption coefficients
(in sq. cm/gram) are roughly (according to Heinrich):

Ni-L in Ni: 1800
Ni-K in Ni: 60
Ni-L in O: 6785
Ni-L in Co: 12650
O-K in O: 1181
O-K in C: 11575

If you use Ni-K and Ni-L ratios from an Ni thin film as a calibration,
you most certainly must consider absorption. The advantage of the
particles of CoO is that you can see the size of the particle and select
a small one where you can estimate the escape distance. The
disadvantage, as I alluded to before, is that I am not convinced that every
particle has exactly the same composition, or, for that matter, that they
are not affected by the electron beam.

I still believe that there is much to be said for studying the Bremmstrahlung
shape, as was proposed in the mid '80's by the people in Glasgow
and others, and on which I experimented and published some
examples. This technique takes account of both the self absorption
in the sample and absorption in any windows, coatings or rogue
layers. It is even able to perform an absorption correction years
later, for it only uses data stored in the spectrum and requires no
other input.

I had it programmed on an old Data General Nova which went
the way of all old clunkers, and I have never ported it to a current
machine. I was able to show that I could get useful analyses from
Ni-20%Al up to a micron thick - the thin-film criterion broke down
for this material at about 10 nm, always assuming you know the
thickness correctly.

I did some experiments with MgO, and was able to obtain moderately
consistent Mg-O ratios over good ranges of sample thickness, but
I was limited in those experiments by the poor low-energy pulse
pile-up rejection of the earlier UTW/windowless detectors. I have
not seriously pursued this with newer detectors.

Tony.

*********************************************************
* Anthony J. Garratt-Reed *
* Massachusetts Institute of Technology, Rm. 13-1027 *
* Cambridge, MA 02139, USA *
* *
* Ph: 617-253-4622 *
* Fax: 617-258-6478 *
*********************************************************