NaI/Tl Gamma spectroscopy - New kit with new sources.

[Seoul_lasers]

I have been on this topic for a while now so please bear with me. I do amature gamma spectroscopy among many other things as a hobby... I am one of these people who seldom finds anything boring/tiresome. Late last year, (Spring of 2019) I finally acquired a very well built Scionix 1.5" dia NaI/Tl detector along with a GS-USB-PRO from Gamma spectacular in Australia. For those of you who are not familiar with this product it has a built in ADC which samples incoming pulses as audio.
The binning of the pulses in done in PRA, Theremino MCA or Berqeulmoni (Japanese MCA software).

Along with equipment I got 2 very strong calibrated sources ( a 10uCi Mn54 disk source as well as a 5uCi Cs137 source). The Cs137 source does nearly 200KCPM on my LND7317 geiger tube. It is highly active! The 2 sources combine to nearly 365KCPM ~110uSv/hr Hence to say, they are both stored in a Lead isotope storage container. They are both considered mid-energy gamma emitters and require much paperwork to obtain in Canada.


Anyways... Here are some pictures of my setup. I am in the process of building a spectroscopy lead pig to shield the background. 0.25" Inner copper layer, followed by Cadmium foil, the followed by ~2-3" of Pb. The shield is meant to block the ambient background as much as possible to have only the measured sample show in the collected output spectrum.

All my spectra taken thus far has been at 550V over 2-4hours periods. Stronger sources require a few seconds up to 5 minutes to produce a useful spectrum. High count rate samples are NOT something you want when doing spectroscopy as they cause pulse pileups.

An added benefit of the GS-USB-PRO is the ability for it to hook up to an iPhone or Android phone to do portable Isotope identification.





Picture insert here




I had posted my pictures here but there seems to be an issue with the server.......

 

[paul1598419]

Cool! I had wondered if you had still been working on gamma ray spectroscopy. I look forward to your photos of the new detector.

 

[RedCowboy]

Would either of your sources be energetic enough to drive a layer of phosphor in order to produce a visible glow ?
I am also interested in seeing your device, it sounds interesting.

 

[Seoul_lasers]

Here are the pics..

Lead pig for the Cs137 and Mn54 (mid-energy gamma storage container) Knocks ~60% of the emissions down.
Still ~100KPM at the surface of the pig.




Exposed Caesium 137 source (toasty warm)





Gamma spectrometer running at 651V.. Note the SHV HV connector on the right is in a single wire mode here. The output is decoupled into an ADC inside the spectrometer.






This is the sweet spot for my probe (550V) - This is the most linear of the voltage settings and produces best results with 10x gain in software. (pictures of incoming pulses coming up)



A 0.3uCi (Cs137 source) - Front of 1.5" Scionix NaI/Tl Scintillation detector - detector set at 550V input gain at 10x



Here is the spectrum generated in ~10 minutes. Note the Gamma peak at 662KeV is due to Beta decay of Ba137m ( m being metasable) ... The 32keV X-ray is due to the two K α-lines of Ba. The compton scatter peak, plateau and edges are also quite visible in the graph.
Resolution is about 6.5% FWHM here, pretty much maximum as you'll get from a NaI/Tl detector.



Scionix NaI/Tl detector with a T bias splitter/filter going into a high resolution audio recorder. Pulses are stretched out between 50-100uSec so they can be processed by a desktop sound card.



This spectrum was generated by a 0.2uCi Ra226 sample ( a watch hand) after about 2 hrs.
The collection of pulses to generate this histogram is around 13.8 million pulses total. There are probably 100x this many that were filtered out. Note the Pb214 peaks and the very prominent 609.3KeV Bi214 peak. This is common to U series decay which Ra226 is part of. The smaller 186.1KeV peak belongs to our target isotope ( Radium 226) The majority of Radium's energy is ejected as alpha particles (He4 nuclei)
while decaying into Ra222 through alpha decay.

This spectrum was captured as a .CSV and then brought into numbers for MacOS 10.15.2




Portable iPhone Gamma spectroscopy. No computer, no problem.
This time we are dealing with Uranium Dioxide ( note the 609.3KeV - Uranium decay chain)
Also can see extra spikes near where Ra226 is.. These belong to numerous U238 and U235 gamma energies.
The iPhone can just barely resolve the peaks due to the crappy ADC sampling rate. Uraninite sample used to generate this is around 27.8KCPS and is only 2cmx1.5cmx1.23cm --tiny but screaming hot sample from a Uranium hotspot in B.C. --or so I've been told.

 

[Seoul_lasers]

Here is a discrimination of Cs137 and Mn54 being done off an iPhone 6S+ Peaks are resolved well 661.7KeV Cs137 and 834.8KeV for the Mn54. Total time for resolution was only 2.5 minutes. Very useful.




A smallish 10Ah Miniso battery bank powers the spectrometer for portable use. In this configuration the spectrometer could run for 62.5hrs without needing a charge.

 

[Seoul_lasers]

Finally got around to adding a digital count rate meter to my Bicron Surveyor MS. I have 2 digital outputs from the meter.
One is a TTL logic pulse (+) and the other (inverted) goes into a PIC24 based pulse counter (yocto PMW-Rx v2) that can stream
over a wireless hub or directly onto a phone or tablet through a Micro USB - OTG connector .

This digital count rate meter is capable of ~3.9MCPM / 65,000CPS

 

[Seoul_lasers]

So, I finally moved into the realm of high speed ( non-soundcard / DIY) Gamma spectroscopy earlier last week. The method was to mate together my Gamma spectacular and a high speed ADC+FPGA + ARM 7v to pull pulses from my NaI/Tl detector at roughly 35nS per sample.
So, you say .. why would you want to do that? While soundcard gamma spectroscopy is certainly great, it takes ~650x as long to pull the same amount of pulses per second as you can with a dedicated high speed ADC and FPGA. The device to make this high speed data acquisition a reality is known as the RedPitaya 125-14 by STEM labs in Slovenia.
The acquisition program (actually 2) are written by members of the RedPitaya community forums.
They are both in a state of WIP (work in progress).

This spectrum was done using 1024 channels....

Professional gamma spectrometers are able to select from a variety of channels in powers of 2.
512, 1024, 2048, 4096, 8192, 16384..

The higher the resolution of the scintillator material the more bins are useful for resolving energies in addition the specific radionuclide being examined.

NaI/Tl (6.5% FWHM) is not high resolution by today's standards.
LaBr3:Ce, CeBr, SrI2:Eu and YGAG are newer scintillators that approach HPGe resolution while being highly portable.




Here is the signal acquisition board Capable of 125Msps @ 14bits
Effective sampling is 40Msps/16bits



High voltage source (Gamma spectular) + Bias T splitter and the RedPitaya by STEM labs



Splitter (Bias T)



~12g of analytical grade ThO2 --- yup its quite a good sample to test detector resolution. Very mildly radioactive
but extremely toxic compound. Activity is around ~50KBq total... I suspect the sample might be slightly more than 12g.
Dose rate is hard to calculate. Gamma ray emissions aren't specifically from Th232. I suspect ~80uSv/hr is more realistic from built up daughter products. The highest energy gamma signature in the sample is from Tl208 which is at 2.614MeV. Th232 is specifically an alpha emitter at 3.98MeV.









Here is the exported spectrum rendered in InterSpec.. All radionuclides are properly ID'd here.