Separating simultaneous pulses???

[tim.hbn]

Hi Everyone

Can anyone here explain to me how gamma spectrometers manage to separate simultaneous pulses?

As an example, I have read that both the 201.83keV and 306.78keV pulses of 176Lu always happen simultaneusly, thus creating a small 508.61keV "sum peak" on the 176Lu spectrum. How is it that the sum peak is not much taller and how do gamma spectrometers manage to extract the 201.83keV and 306.78keV pulses from what I would expect would look like a single 508.61keV pulse?

Thank you very much.

Kind regards

Tim

[Sesselmann]

Tim,

The simple answer is they can't..

Especially with relatively slow sampling as used by most of the people on this forum, it is impossible to separate pulses that occur within the dead time of around 100 µs.

The software treats this as pulse pile up and simply drops the pulse, or as you realised, counts it as a taller pulse.

Theoretical detective work..

Steven.

[tim.hbn]

Hi Steven

Thank you very much for your reply.

If you look at Christoph Denk's 176Lu spectrum in the following link, you can see that his gamma spectrometer succeded in separating the 201.83keV and 306.78keV pulses.

https://www.gammaspectacular.com/blue/lu176-spectrum

How was this possible?

Thank you very much.

Kind regards

Tim

[GigaBecquerel]

It's possible, because while the two gammas are emitted within a few ns of each other they both fly in random directions, and since your gammaspec (likely) won't see every gamma coming off the source there's a very good chance you'll only see one of the two photons per decay.
If you had a (close to) 4pi detector the summing peak at 508 will be a lot stronger, but the less sphere surface your detector covers the less chance there is for a summing peak.
But it's true what Steven says, if both photons hit your detector there is no way to separate them.

[tim.hbn]

Hi GigaBecquerel

Thank you very much for your reply. You have solved the mystery for me. What you say makes perfect sense.

However, I have one question for you. What is a "4pi" detector?

Thank you very much.

Kind regards

Tim

[GigaBecquerel]

The surface of a sphere is 4*pi*r².
A detector covers a certain part of the sphere, and for lazyness it is pretty common to refer to that number just as a fraction of the whole sphere.
A 4 pi detector sees everything that comes off your sample, a 2 pi detector sees half of it etc.

[tim.hbn]

Hi GigaBecquerel

Thank you very much for your explanation. You are very good at explaining things.

So, it seems to me that a 4pi detector is a somewhat theoretical concept. I am guessing that in order to create one you would have to hollow out a large scintilator crystal and place the sample right inside it.

However, as I expect you will agree, it is much more useful to be able to see the 201.83keV and 306.78keV peaks of 176Lu than it is to see the sum peak. Can you think of anything which would make it worth building a 4pi detector?

Thank you very much.

Kind regards

Tim

[GigaBecquerel]

True 4 pi detectors are actually a thing!
They have many different uses, but it's true that you have to watch out when using them for spectroscopy. Most of the time they're just used for counting, to detect very low level activity, or as a compton shield, or cosmic veto.
But segmented 4 pi detectors are more common, where you have many detectors all pointing towards one point, to be abe to separate individidual photons or whatever particles. Common examples would be the euroball detector, Griffin and many more, but the most well known will probably be the CERN ATLAS, CMS and ALICE.
PETs for medical use also are close to 4 pi, and they function only because in positron annihilation two photons are emitted at the exact same time, looking for just those coincidence counts!

[tim.hbn]

Hi GigaBecquerel

Thank you very much for another brilliant explanation. :-)

I am therefore guessing that with a segmented 4p detector, you get the best of both worlds and the smaller each segment with a higher number of segments, the better the resolution.

Thank you very much.

Kind regards

Tim