[NewCandle] los artifactos

[John Winterflood]

Yo Nick,

It just occurred to me what your Y peak is!

It is the same "pile-up" effect occurring, but instead of two Al pulses 
arriving within the pulse discriminators window, two pulses from 
different elements arrive arrive at the same time so that their energies 
*sum* to give the Y energy. In this case the prime contenders are Al 
(1487) and 0 (524) which appearing together will sum to give 2011 which 
is close enough to be confused with Y's 1923 (100%) and 1996 (57%). (For 
those who are interested I have been looking at the charts given at 
http://tinyurl.com/5oorjr and http://tinyurl.com/5zpt5c - nothing 
special about them, just the first that Google came across).

Nick Reiter wrote:
> ..
> even a bona fide Ag signal (from looking at 200A of Ag
> deposited on Al foil)
Something that bothers me about attempting to look at such *thin* layers 
is that this is not the volume that is primarily sampled by the high 
energy electrons. If you can find a diagram of the sampled volume you 
will see that it is very pear shaped going quite deep (several microns) 
with scarcely any x-ray sampling occurring on the surface (the stalk of 
the pear). I tried to find a diagram of this pear shaped volume on the 
web but was unable to in the time I gave it. I did come across some 
useful information however - firstly from someone's power-point talk: 
"X-ray analysis pitfalls: 1) EDS is *not* a surface technique – 
beam-sample interaction volume increases with beam energy". And I found 
some depths quoted: "Low energy provides surface imaging with a thin 
interaction volume of <0.2 µm at 5 kV, compared to an interaction volume 
depth of 1.5 to 2.0 µm at 15 kV". You will note that your silver layer 
was only 0.02 µm thick and you are running at 30kV - which probably 
gives you an interaction depth of 3 to 4 µm, and a pear-shaped (or 
approximately spherical) volume centered at that depth. So you are 
sampling many hundreds, if not thousands, of times more volume of 
aluminum than silver. With a detection limit of order 1% it is a wonder 
you can see your silver layer at all!
>  ... Apologies to all for the runaround.
>   
No apology needed, rather thanks is are order for presenting some 
interesting data.

Earlier Keith wrote:
> .. As far as the list,
> I suppose I'm partly responsible ..
Gosh I didn't mean to criticize this list, rather I intended only to 
show my appreciation for Nick's large contribution!

This is a great list - low traffic and high signal to noise when there 
is some traffic.

J.