Lead Castle Improvements and Low Level Measurements

[GigaBecquerel]

Hello Guys!

Long time no see. Things have changed and thanks to a few friends, Luuk among them, my gamma spectrometer got a real upgrade!
From Luuks retirement sale I got a perfect 3x3" NaI Detector, complete assembly with PMT and Crystal. He tested it to be 5.8% @ 662 keV, and who am I to doubt his measurements ;-)
Kidding aside, I am very happy with the detector! This will definitely get a lot of use in the future. But a good detector needs a good shielding setup to make full use of it, this is what I'm writing about here.
My lead castle consists of 5 cm thick lead bricks on the walls, and 20 mm lead plates on the top and bottom. It's all a bit thinner than I'd like, but all in all this already comes close to 200 kg and I don't want to load the floor much beyond that.
Inside the lead is a 3 mm thick copper layer, this is there to absorb lead XRF and 47 keV from Pb-210, unfortunately my lead is not well aged low background. This copper layer holds a "Box" made from 10 mm thick PMMA, which absorbs any betas coming off my samples to prevent them from producing any Bremsstrahlung or X-rays in my metals. The Box is made from Offcuts, so it cost me maybe 5€ and it's such a huge upgrade! If you want to improve your shielding I highly recommend you give something like this a try, it really can't be beat in terms of cost versus performance.
Inside the PMMA is the actual detector and volume for samples to be measured, with the detector already taking up a good amount of that space. But with something this sensitive you don't need big stuff to get big results anyways.
My overall background is 13 counts per second, with most of that coming from a 0 to 3 MeV continuum, presumably compton scatters and bremsstrahlung from whatever is out there. From this a few peaks arise, mainly 511 keV (0.10 cps) and 1460 keV (0.15 cps), those I cannot get rid of with any reasonable effort. Maybe I should go on a low potassium diet to improve those numbers further ;-)
Beyond 1.5 MeV I get a few peaks of gammas that simply make it through the lead, 5 and 2 cm simply aren't that much, 10 or more would definitely give an improvement, but again, floor loads. At least their spectra agree with them having passed through 5 cm of lead, and not originating from inside the shield.
If you want to reduce background you have to make sure your chamber is clean, especially if you measure loose samples or anything that emits radon. A wipedown with a wet towel already helps wonders! This is another big advantage of the PMMA, it gives big, smooth surface that can easily be cleaned.

I'm sorry for posting links instead of uploading the images directly, but I've tried that and I always get an error, so here you go.

https://assets.chaos.social/media_attac ... 115f1e.jpg
https://assets.chaos.social/media_attac ... e28c32.jpg
^ Building the Castle

https://assets.chaos.social/media_attac ... 682afe.jpg
https://assets.chaos.social/media_attac ... 179d37.jpg
https://assets.chaos.social/media_attac ... 780e3a.jpg
^ Adding Copper

https://assets.chaos.social/media_attac ... 519b7a.jpg
^The finished Assembly with detector

All of this is nice, but what does this actually mean in terms of being able to see what's out there?
To demonstrate this I have a fun experiment. Some Phosphate Removers for Ponds are based on Lanthanum chloride, Lanthanum Phosphate is near insoluble in water so it just drops out of solution, leaving back the chloride anions.
Lanthanum is slightly active due to it's ¹³⁸La content, so let's see what we can see in these phosphate removers! I bought two bottles of 100 ml each that say they contain 12% LaCl3 * 7 H2O. At that concentration the solutions density can be assumed as 1, so that makes 12 g of the heptahydrate per bottle. The molar weight of that is 371.37 g/mol, giving me roughly 32 mmol per bottle. 32 mmol of Lanthanum are 4.4 g, the natural concentration of ¹³⁸La is 0.089% so we have just about 4 mg. The specific activity is 941 Bq/g, giving me ~3.75 Bq per bottle or just over 7 Bq in total. The fun thing is that ¹³⁸La has a relatively low gamma emission probability, 0.656 for the 1436 keV gamma and only 0.344 for the 789 keV gamma, giving me 5 and 2.4 gammas per second for these lines. Multiply that with the counting efficiency of my setup and you see where this is going, I get an extremely low count rate.

https://assets.chaos.social/media_attac ... 41a07b.jpg
https://assets.chaos.social/media_attac ... f7f021.jpg

After ~2.5 hours the 789 keV peak is well visible with 0.09 cps, as well as the 1436 keV peak, but here it's hard to say anything about the actual efficiency as this strongly overlaps with 1460 keV from potassium.
I'd call that a successful low level measurement!

Lukas

[Sesselmann]

Lukas,

Thanks for posting, that's probably the most elaborate lead castle I have ever seen on this forum, congratulations and well done.

I imagine it was not cheap, interlocking lead bricks and thick copper sheets $$$

You are lifting the bar and setting a new standard 👍

Steven

[GigaBecquerel]

Good Morning Steven,

Thank you!
I got very lucky with this castle, and I have been collecting materials for it for the last 5-10 years. The lead is from a nuclear medicine department that closed down, I bought it for around 1€ per kg. The copper is cut from offcuts by a company that does contract lasercutting and was already payed for by a customer... you have to know someone who knows someone, you know how it works. The PMMA is from a local plastics supplier, they sell their scraps for 7€ / kg and I used the lasercutter to make it. All in all I think this is 300€ in materials, plus a weekend of work spread over many years.

With patience and a few connections everything is possible.
Lukas