Scintillation Detectors:Gamma-Ray Spectra

Finally, you can see here the reason that
we go to all this trouble to make and use semiconducting detectors. The first picture is of a cobalt-60 spectrum
taken with a sodium iodide detector, and the second here is a spectrum taken with a high
purity germanium. Notice the resolution of the two peaks in
the high purity germanium. What a difference thirty times difference
in resolution can do! Okay, I just saw one of you yawn and say,
“So what?” Look at this. This is why we pay the money. What sodium iodide smears out beyond recognition,
high purity germanium resolves into individual peaks. We can go and identify minor constituents
in the material that we can only dream of when using scintillators. Germanium detectors come in a variety of shapes
and thicknesses of the p-n junction and of covering materials. Ultra low-energy germanium detectors have
ultra-thin junctions and carbon fiber windows, so they can be used in the sub-keV photon
range. And, as a hint, bring lots of money if you
want one of these. Low-energy germanium detectors can have beryllium
windows, and the rest use aluminum. Coaxial detectors give a lot of volume, but
have a relatively large dead region at the surface, so they are only good for photons
50keV and above. Well-type detectors are built with an inner
well that holds small samples, so that you can have almost a 4 pi geometry. Here’s the effect of superior resolution
of the high purity germanium compared to sodium iodide for highly enriched uranium, and the
two spectra are superimposed, the high purity germanium in green, and the sodium iodide
in burnt orange. The burnt orange is obviously inferior. Here is another example showing sodium iodide,
high purity germanium, and cad-zinc-telluride, another type of semiconducting detector. The resolution of the cad-zinc-telluride is
not as good as high purity germanium, but it’s still far superior to sodium iodide. I think that by looking at the three spectra
here, it’s easy to understand that resolution is very useful in complex spectra, like uranium
and plutonium. A very important feature here is the 186keV
peak from U-235. It is this peak that we use as a quantitative
measure of the U-235 in the sample. Also notice the uranium-characteristic x-rays
at around 100keV. We can hypothesize that the sample we measured
here is HEU, based on the height of the 186keV peak to the uranium x-rays. The source material for these three spectra
is freshly-made HEU. How can we tell that? By the ratio of the 186 line to the uranium
x-rays, as we said, but also by what’s not here. Above 250keV there are none of the peaks of
gammas emitted by the in-growth of radioactive daughters. Well, what do I mean by all this talk of radioactive
daughters? Here’s the simplified decay scheme for the
thorium-232 series. Many of these daughters emit gammas, as we’ll
see in a minute. This series is also known as the 4n series,
because, if you look carefully, all of the daughters’ mass numbers are exactly divisible
by 4. We will skip the 4n+1 series, because we don’t
need it right now, and look at the 4n+2 series, or uranium-238. Again, as the daughters grow in, they will
emit gammas. And, finally, here is the 4n+3 series, or
the uranium-235 series, and it is these daughters that we didn’t see in our spectrum. If the sample were LEU, we would have a Compton
continuum in that area from the 1001keV peak from U-238, which is actually from the daughters,
but who’s counting? Since detectors are much more efficient for
counting low-energy gammas than high-energy ones, it’s very common for us to measure
the enrichment of a sample by measuring the relative strength of the 186keV peak to the
Compton continuum above the 186keV. Now, let’s look at a spectrum that is not
quite so pretty and see where this Compton-to-186 line ratio would lead us astray. If we look at our 186 line, we see that in
this spectrum it sits on top of a large Compton continuum generated by high-energy photons. The lead-212 gamma is a clue that something
is not right. This is the daughter at the bottom of the
thorium chain, not the U-235 chain or U-238 chain. If we zoom out, we now see that there are
peaks from in-growth daughters above 250keV. This sample is a mixture of uranium and thorium. The thorium could just be there because both
elements were together for some reason, or it could be from U-236 that alpha-decays to
thorium. The U-236 is from reactor return uranium,
that is, stuff that’s been irradiated before and being put back in fuel. These daughters will contaminate hadthe Compton
above 185 peak, and it would lead to an erroneous result. In this case, us thinking that the sample
had a lower enrichment than it actually has. What we could do? We could take a broad-energy spectrum, and
then the 238 peak could be measured directly, and we could ratio the 186keV peak to the
1001keV peak, although this is actually not a common thing to do, since you have to count
for a long time to get food statistics in the 1001keV peak. Okay, what does LEU’s spectrum look like? Here you can see the 238 gammas at high energy
and how they’ve forced the 235 gammas to ride on their Compton continuum, which make
the Compton-186 line ratio useful to assess enrichment, but also realize that this ratio
will change with time as the daughters grow in.

You may also like...

3 Responses

  1. Frank Cass says:

    nice clear presentation of why HPge is good idea

  2. Дмитрий Бурбовский says:


  3. pushkar kushwah says:

    sir send some vedioes on following
    1. Detector electronic and pulse processing
    2. pulse counting system
    3.pulse height analysis
    4.pulse timing ,pulse shape
    5. pulse shape discrimination

Leave a Reply

Your email address will not be published. Required fields are marked *