NaI scintillator resolution and rejuvenation

[richnormand]

I am having a difficult time resolving higher energy peaks with my Harshaw 3” by 1” NaI scintillator. For example I barely distinguish the double peaks of Co around 1.25MeV.

It is read by a Canberra series 3 that performs well when fed from a variable height pulse generator or with a Bicron x-ray detector. I have to conclude that it is the crystal or the PMT.

The crystal has a few cracks in it but it is still transparent with only a hint of yellow. Some places offer to “rejuvenate” old NaI crystals. Considering it is hydroscopic I would assume they simply leave it in an oven under low humidity and high temperature?

Any ideas from the gurus here?

Cheers and thanks

[Frank Sanns]

If the crystal is cracked you will loose your resolution forever. It will still detect but will no longer resolve.

Heating an already cracked crystal a great way to have it crack some more. I would not be too kean on heating up the crystal even slowly as the internal stresses from cracking will build up higher and the chance of more cracking is too great. Pulling a vacuum on the NaI will get the moisture out and I have even heard that it will remove some of the yellowness if you pump it down deep enough leave it under vacuum for half a day (or over night).

Frank S.

[richnormand]

I have a Canberra Series 3 this is a full MCA with all the trimmings as well as a Canberra 1705 signal processor. Works well at high resolution with other detectors. Looks like it its the Harshaw NaI/PMT that is doing it. The crystal has multiple cracks at the bottom edge. I suspect multiple light scattering could be the culprit of built-in stress from the fracture. This was an e-bay purchase a while back and I only opened the housing recently to try to find out why the resolution is not that good.

How easy is it to cut and polish NaI? I have a very good region in the middle of the crystal (its a 3" diad by 2" thick)

[Richard Hull]

Note* If you are placing the source face-on to the PMT it might not do as well as a 2"! You only have 1" of stopping power of crystal from the front which is not good. Place the source to the side (shining in through the 3" section and the crystal) and see if you improve. 1" thick crystals are crap for high energy gamma spectro work. The absolute best gamma spectro NaI xtal is a 3"X3".

Finally, any cracks in any scintillator are just about a show stopper, as mentioned above.

Richard Hull

[Frank Sanns]

I had a thought about the cracked NaI crystal. It may be possible to repair it sufficiently to restore the original uncracked performance. From another current post ( viewtopic.php?f=13&t=5671#p34115 ) about coupling scintillators to PMTs, the role of optical coupling fluids is discussed. The thought is to repair the cracks with an optical coupling adhesive with the same refractive index as the xtal. I would imagine you could put the adhesive over the outside of the xtal, pull a modest vacuum, then release it. The adhesive should be sucked into the cracks and fill them. Wipe the residual adhesive from the xtal faces and let it cure. Should work. This is similar to the process used to repair small cracks in car windshields (windscreens).

Frank S.

[richnormand]

Good idea. Jon mentionned using high weight silicone oil (as used in RC car differentials) as coupling fluid for his plastic muon detector. I'll look up the index of refraction of lucite/plastic compared to NaI. If close this might work and seep into the cracks? Nothing much to loose unless the oil damages the crystal

[richnormand]

You are right.
I usually place the source about 5cm away on the side to get the full 3" xtal cross section. The front has a stainless steel plug on it to position the assembly. This is the detector on the far right of my pix in the muon thread.

[JohnCuthbert]

Does anyone know if you can repair a cracked NaI crystal this way?
If you can then, so far as I can see, with a good refractive index match you could "glue" together powdered NaI with a suitable resin and make any size detector you wnted.
This would give you most of the sensitivity and resolution of a single NaI crystal and a lot of the convenience of a plastic scintilator.
This seems too good to be true, so I guess there's a mistake somewhere.

[Frank Sanns]

I too would like to hear from Carl and some of the other spectroscopy experts here but here are some of my thoughts.

For gross count detection only, I would expect that the efficiency would be approximately the weighted average of the amount of NaI:dilluent ratio. The dilluent could be a matched refractive index glue or liquid or scintillator containing organic resin. So if you had 80% NaI and 20% non scintillating binder resin, you might expect 80% of the NaI effeciency.

The efficiency may be less than that because NaI(Tl) is essentially transparent to its scintillation wavelength. This is not the case with plastic scintillators.

For spectroscopy, my knee jerk reaction is that it would be very poor. It would be one thing to repair a single crack with a thin optical interface of fluid and something else entirely for 10^23 cracks. Then you have the problem of "wetting" the entire surface of the NaI powder. Any air void or places that the resin in not in atomic contact with the NaI would yield a path variation that would kill the resolution. Again, I am not the expert here and I would really like to hear from the others. No matter, it would still be an intresting experiment to do as I would really be curious of the spectroscopy result.

Frank S.

[Carl Willis]

The idea of putting the crystal in an index-matched liquid (one that is dry and otherwise won't attack the material) is interesting. For small cracks I don't think you'd see much reduction in resolution compared with the undamaged crystal if this method were used.

But the suggestion gives rise to a more severe idea that I don't think would work well for resolution: what if you took a bunch of tiny grains or crystals of NaI(Tl) such as one could grow at home by evaporation, and stuff them in a jar of liquid with refractive index of 1.85 or so? I do think this would reward you with a monster scintillator, but electronic equilibrium within the NaI microcrystals is likely to suffer for high-energy gammas. Electrons from photoelectric and Compton processes will lose some energy traversing nearby inert material, with the end result that you have a broad full-energy peak. However, this idea has merit if you are looking for efficiency.

-Carl

[Q]

hmmm, a home grown scintillation crystal!
very interesting! since i'm not at the sprctroscopy stage yet, i may give that a try.
one question that comes to mind though, aren't NaI (and other crystals as well) usually doped with some other metal like titanium? how would one go about doing this in a home lab situation?

Q

[Richard Hester]

NaI crystals are doped with Thallium, which is exceedingly poisonous - not something I'd want around the house. Back when I was little (a long, long time ago), Thallium Sulfate was used as an ant poison - it's colorless, odorless and tasteless. A couple of kids died by Thallium poisoning through skin absorbtion playing with the "empty" tins. The stuff was withdrawn from the market. The doping percentage is pretty small for a NaI:Tl crystal, but I wouldn't want the stuff around unless I had a professional quality lab setup.

[Jon Rosenstiel]

Here is my $ .02 on this subject....
My original intent was to build a detector suitable for gamma spectroscopy using one of Frank S’s BGO xtals. Then, this thread came up and got me thinking about a multiple crystal detector. Being I had a small supply of the BGO crystals I decided to build and test a multiple crystal detector. Two goals: One, would multiple crystals optically coupled act as one large crystal? And two, to hopefully shed some light on how cracked and repaired crystals behave.

I coupled six of the 6 x 12 x 30mm BGO crystals (Frank S. BGO xtals) to the face of a 2" pmt. The xtals were positioned with their 6 mm edge against the pmt face. A small amount of silicone fluid (10,000 weight RC car differential fluid) was used between the xtals and the pmt and between the xtals 12 mm faces. (See upper photo)

Using my normal gamma spectroscopy setup (Canberra series 35+ mca) I did a comparison between the multi-xtal BGO detector and my Bicron 2M2 (2” x 2") NaI detector using the 662 keV gamma from a 10 μc Cs-137 source disc. (See lower photo)

Results: The resolution of the multi-xtal BGO detector was in the crapper! My Bicron NaI detector has a resolution of 7.3% FWHM at 662 keV. The multi-xtal BGO detector’s resolution ran from 14.7% to 17.4% FWHM at 662 keV.(For reference, a detector using a single 6 x 12 x 30mm BGO crystal had a resolution of 10.4%). With the detector faced up to the source disc the resolution was14.7%. With a source to detector distance of 5.6 cm the resolution was 17.4%. Intermediate source to detector distances gave intermediate resolution values. Also, if the source disc was moved laterally (with same source to detector distance) the resolution would change.

With five crystal to crystal interfaces this was an extreme example as far as cracked crystals go, but I do believe it shows what to expect with a cracked and/or a cracked and repaired crystal scintillator.

1. Degraded resolution.
2. Inconsistent resolution. (Dependant on source to detector geometry).

Same basic conclusion as mentioned earlier in this thread, a cracked crystal is curtains for spectrometry, but not a big deal if all you want to do is detect radiation. Take it U rock hounding, what do you care if you drop it, it’s already cracked!

Jon Rosenstiel

[Q]

thanks richard,
oops, thallium... not titanium. i knew it was one of those "t" elements. to many days with not enough sleep.
anyway, yeah, thallium isn't a substance that i really want to mess with in the current lab setup. but do i understand correctly that one could dope a NaI crystal simply by adding a tl salt to the solution?

Q

[Richard Hull]

Cracked crystals are goners. Forget cracked crystal as useful items. They are forever degraded and pretty worthless.

This is why it is absolutely imperative to have an intellegent and cogent seller supply, in writing, that any PMT with xtal is in flawless condition prior to purchase with a full money back on defective items.

Of course on e-bay or in a surplus situation this is going to happen about .0000002345% of the time. Earl or Elmo don't know nuthin' about no things like that............I got this thing and I want money and you ain't gittin' yers back once you got my thing.

Scintillation crystal are far too easy to damage through mishandling or temperature gradients. Most are surplused due to cracked or otherwise horribly degraded crystals, but they sure do look purdy from the outside.

Richard Hull

[richnormand]

Interesting.

I used one of my $9 e-bay 2” PMTs for this test, so results may vary…!

I first used one of the BGO with a polished 12mm face on the PMT and a Cs source at 5cm. Resolution was a bit better than 11%. This I think is consistent with what is to be expected. Also in agreement with this measurement,
viewtopic.php?f=13&t=5677#p34121
First pix.

Then considering the comments here

viewtopic.php?f=13&t=5672#p38002
Second pix.

I wondered about polishing the other two crystals since it takes me about half a day to do a good job with diamond paste for all surfaces on each one.

Well I tried it anyway. With three crystals with RC diff fluid coupling between the xtals and the PMT ( thanks for the tip Jon about RC diff silicone stuff), I got about 11% resolution but a lot faster acquisition. I’ll repeat the whole process in a more systematic way to check these results in the next few days.

Here is what I got with three xtals:

See pix #3.

with that geometry that was put in a cardboard box (with light leaks too) for a quick test. The Al foil is also seen :

Pix #4


Comments welcome but it seems that with a good optical interfaces the resolution is not degraded as much as when adjacent surfaces are of poor optical quality.

[Jon Rosenstiel]

Very interesting, Rich. You've done an awesome job of polishing those xtals. How's the fingers feeling?

I did a real quick polish test with my buffer, and it appears that it will do the job, but those crystals are a bear to hold onto. To keep flying crystals to a minimum I'll need to make some type of holder.

Keep us posted.

Jon Rosenstiel

[Frank Sanns]

Wow, you sure made the xtals sparkle. How would you like to do the 39 that I am tiling my 5" PMT???? I think not.

I am curious how the interface LOOKS with the coupling fluid between the xtals. Can you see the interface at all or is essentially totally gone? If the refractive index is close to the substrate, the interface will dissapear and your three xtals will look like one big one. It is this condition that really interests me.

Good work Rich! When I get my interface completed in the next day or so, I will run my MCA / BGO monstor xtal to see what the limit of resolution will be.

Frank S.

[richnormand]

Here is a photo looking through the side. The interfaces do vanish.

I put a microscope slide at the entrance and exit surfaces since the RC silicone fluid was creating distortions and I did not want to remove the whole assembly for a cleaning. The black bars in the back were traced in pencil on a ¼” sheet placed about 4 cm away.
The coupling fluid does effectively remove the slight haze and orange peel effect left by my polishing job.

(Edit:) Looks OK in the visible but it is at the UV emission wavelength that the interfaces and coupling fluid matching would be important.

I started with 6um diamond paste and moved immediately to the 0.5 um paste and used a polishing pad that was a bit too soft and thick. Not the best way to do it but I was afraid of breaking/dropping the xtals so I skimped on the polishing time as much as I could. Also my finger tips still hurt from the sharp edges holding these suckers! If I was to do a larger quantity I would invest in creating a suitable jig and move to a motorized lapping polisher. That would also preserve the surface flatness better although this does not look as a problem.

[HighVoltageFox]

I don't know if this has been said yet but anyway; the solution to the water absorbing crystal is a vacuum oven, commonly found on ebay and labX on the cheap. Of course those of you with extra vacuum fittings can just throw a small vacuum chamber together and just pump it down and heat it or bake it out with the crystals inside, quick cheap vacuum oven.
Oh, yeah and I have some experince in glass blowing and generally broken glass is a lost cause unless you want to melt it down.
,Andrew

[richnormand]

here is the result with a fourth xtal. Cs source at about 8 cm away. PMT is a 4883 RCA from e-bay

[Jon Rosenstiel]

That looks pretty darn good, Rich. Does the resolution remain the same when you change the source to detector distance? Or when you move the source laterally?

Jon Rosenstiel

[Frank Sanns]

Actually, I think the most accurate word is deliquescent.


Frank S.

[richnormand]

Actually, not too bad.

I ran a few tests today. The MCA (Canberra series 30) was set to count 500 at the peak of the Cs source spectrum. At close distance, and sideways depending on the 4 xtal orientations.

Makes me think a 5th xtal to fill the footpath of a 2” PMT may be worth it in the geometry previously described.

Comments are welcome.

[Jon Rosenstiel]

It appears that polishing the crystals help, as your resolution numbers are a lot better than what I got using unpolished xtals. It's still not what one could call a "lab quality" instrument, but not bad, not bad at all.

Jon Rosenstiel