Well, based on my experience, the XRF appears to be identifying
materials based on a certain “window” peering into the metal. it can
vary, but using the attenuation and filtration of say, a 5 to 30 Kv
pulsed x-ray spread (the actually energy of the x-ray) I seem to be
seeing a conical shaped window, up to about 5 mm thick, at about a
55 degree angle of dispersion. I have not actually bothered to set up
a test station and measure this, because I don’t really care, but
that is what it looks li,e.
plating can confuse the unit, on occasion, as these thin coatings
act as, I think, fourier filters, which can drop out every other
backscatter wavelength. confusing the instrument. I had this happen
on a piece of plated steel, it reported it as gold over chrome, it
wound up being a very thick rhodium coating over platinum over steel
(not something you would see in jewelry, this is out of a high
energy physics piece of equipment, a linear accelerator, weighs about
11,000 pounds, no one is going to walk into your shop with it).
For true determination, we still use backup fire assay with a
reputable assay house. for example, we have one piece of metal that
the portable XRF has identified as 88% tungsten, 11% gold. Normally
we wouldn’t be too concerned, but this block weighs 84 pounds (no I
did not leave out the decimal point) so we are going to get a 2nd
opinion before we celebrate.
however, on straight allows, it is very accurate, +/- 0.02 %,
probably I have that much variation in metal content in my alloys in
jewelry work, simply due to density separation when melting.
The religious charms were scary. With the high cadmium content
(46.0%, not 0.46%), and as poisonous stuff as this metal is (it is a
neurotoxin), the net effect is that the more you rub the charms, the
crazier you are going to get.
Now, if anyone who is reading this wishes to disbelieve, or slay the
messenger, or whatever, well, if that is your reaction, well then,
that is just that, your reaction.
As always, presented here is to be used, or not, at your
own risk.
So, to answer your question directly, I myself have taken a reading
of 46% cadmium on a charm with a hand held portable XRF.
Warm regards
Mark Zirinsky, Denver