Neutron activation with 10 mCi Po + Be
Posted: Mon Mar 27, 2017 12:26 am
While it's a long-term goal of mine, I don’t have the resources or technical know-how to make a fusor yet, but I still wanted to play with a few neutrons and activate some stuff. The best way to do this would be Carl Willis’s AmBe source, but I don’t know where to get a whole bunch of 60+ uCi smoke detectors and I’m not sure on the legality of this anyway.
So I did the next best thing: I leased two 5 mCi P-2042 Po-210 static eliminators and bought a couple of pieces of an old beryllium X-ray window from ebay, which I cut down to size. For people who haven’t seen the P-2042 before, there is a little wire mesh about 2 mm above the gold/polonium surface of the ionizer, so direct contact between the Po/Au and the Be isn’t possible without either violating the law and removing the mesh, or grinding the Be up into powder, which would both be bad ideas.
I taped the Be to the P-2042 so that the whole area of the P-2042’s mesh is covered; hopefully I’m not losing too many alphas and alpha energy to the mesh and the ~2 mm of air. I got some 12”x12”x0.75” HDPE sheets and some smaller 0.75” thick HDPE blocks to moderate the neutrons. Four of the sheets are on the bottom, then there are the sources and the material(s) to be activated surrounded by the smaller blocks, and then five more HDPE sheets are on top. I usually leave things in the neutron oven for several half-lives.
For detection, my mainstay is the GMC-320. It’s awesome - a Geiger counter for only $120 with built-in second-by-second data logging, and the battery lasts for ~4 days on a charge. The data can then be exported as a CSV and analyzed with Excel. I take the stuff I’ve activated, usually in baggies if a powder, open up the GMC, and tape the baggies or foil directly to the tube so that the GM tube is surrounded as much as possible by the (hopefully) activated material.
Then I wrap several ~1.5 mm thick sheets of lead around it to cut down on background, stick it under the bed, and wait for at least six half-lives of the activation product. I take the average of everything past about 4 or 5 half-lives to be background, and subtract that from the number of counts recorded each minute. If the average of the counts for the first half-life is above background in a statistically significant way (preferably p << .01), then I have some good evidence that my neutrons did something.
For things with half-lives above 12 hours (e.g. tungsten, gold), that method won’t work as well. I’ll probably use the old Inspector pancake Geiger counter I picked up for a bargain on ebay. Its main drawback is that it doesn’t connect to any electronic devices, so no data logging. It will count for a set amount of time, though, so I can take periodic 1-hour readings every so often and see if those change over time. I’ve started with tungsten, but got a negative result there.
Using the GMC-320, I can say with a high level of confidence that I’ve activated indium and manganese. A bar of gold and some dysprosium oxide are currently in the neutron oven, and europium oxide is currently taped to the Geiger counter in the lead sandwich under my bed and is being counted. I was, however, not able to detect Mn-56’s 850 keV gamma with the cheap Russian-made gamma spec I have. I’m hoping to have better luck with Au or maybe a second try with In.
Here are my results from the GMC-320 for In (17 g, mostly foil) and Mn (110 g of MnO2). As you can see, the increase above background is not huge but is definitely statistically significant for the first half-life (p = .00003 for In, p < 10^-10 for Mn). Mn is significantly (p < .0001) above background for its second half-life as well, and In is marginally above (p=.04). They both return statistically to background levels by the third half-life.
The GM tube is fairly small and not all that sensitive, so background is only about 14 cpm within the lead sandwich; an increase of 3 cpm is about a 21% increase overall. I’m posting the 15-minute running average of the CPM for each one to smooth out some of the random wiggles. Both graphs are chopped off; I continued the Mn one out to 10 half-lives and the In one to 15; the averages from 5 half-lives on were averaged and made up the background, and nothing like the spike at the beginning of each occurs in the second half.
So, I have a few questions. First, when alpha particles hit beryllium, is the radiation of the neutrons isotropic, or are neutrons emitted in some anisotropic way? Related to that, once the neutrons hit the hydrogens in the moderator a few times and thermalize, what does the distribution of thermal neutrons look like? Is it more or less isotropic with greatest density right next to the source and diminishing in inverse-square fashion after that, or does it look different? If not, is it possible that putting a layer or two of HDPE between the sources and the target is a good idea?
So I did the next best thing: I leased two 5 mCi P-2042 Po-210 static eliminators and bought a couple of pieces of an old beryllium X-ray window from ebay, which I cut down to size. For people who haven’t seen the P-2042 before, there is a little wire mesh about 2 mm above the gold/polonium surface of the ionizer, so direct contact between the Po/Au and the Be isn’t possible without either violating the law and removing the mesh, or grinding the Be up into powder, which would both be bad ideas.
I taped the Be to the P-2042 so that the whole area of the P-2042’s mesh is covered; hopefully I’m not losing too many alphas and alpha energy to the mesh and the ~2 mm of air. I got some 12”x12”x0.75” HDPE sheets and some smaller 0.75” thick HDPE blocks to moderate the neutrons. Four of the sheets are on the bottom, then there are the sources and the material(s) to be activated surrounded by the smaller blocks, and then five more HDPE sheets are on top. I usually leave things in the neutron oven for several half-lives.
For detection, my mainstay is the GMC-320. It’s awesome - a Geiger counter for only $120 with built-in second-by-second data logging, and the battery lasts for ~4 days on a charge. The data can then be exported as a CSV and analyzed with Excel. I take the stuff I’ve activated, usually in baggies if a powder, open up the GMC, and tape the baggies or foil directly to the tube so that the GM tube is surrounded as much as possible by the (hopefully) activated material.
Then I wrap several ~1.5 mm thick sheets of lead around it to cut down on background, stick it under the bed, and wait for at least six half-lives of the activation product. I take the average of everything past about 4 or 5 half-lives to be background, and subtract that from the number of counts recorded each minute. If the average of the counts for the first half-life is above background in a statistically significant way (preferably p << .01), then I have some good evidence that my neutrons did something.
For things with half-lives above 12 hours (e.g. tungsten, gold), that method won’t work as well. I’ll probably use the old Inspector pancake Geiger counter I picked up for a bargain on ebay. Its main drawback is that it doesn’t connect to any electronic devices, so no data logging. It will count for a set amount of time, though, so I can take periodic 1-hour readings every so often and see if those change over time. I’ve started with tungsten, but got a negative result there.
Using the GMC-320, I can say with a high level of confidence that I’ve activated indium and manganese. A bar of gold and some dysprosium oxide are currently in the neutron oven, and europium oxide is currently taped to the Geiger counter in the lead sandwich under my bed and is being counted. I was, however, not able to detect Mn-56’s 850 keV gamma with the cheap Russian-made gamma spec I have. I’m hoping to have better luck with Au or maybe a second try with In.
Here are my results from the GMC-320 for In (17 g, mostly foil) and Mn (110 g of MnO2). As you can see, the increase above background is not huge but is definitely statistically significant for the first half-life (p = .00003 for In, p < 10^-10 for Mn). Mn is significantly (p < .0001) above background for its second half-life as well, and In is marginally above (p=.04). They both return statistically to background levels by the third half-life.
The GM tube is fairly small and not all that sensitive, so background is only about 14 cpm within the lead sandwich; an increase of 3 cpm is about a 21% increase overall. I’m posting the 15-minute running average of the CPM for each one to smooth out some of the random wiggles. Both graphs are chopped off; I continued the Mn one out to 10 half-lives and the In one to 15; the averages from 5 half-lives on were averaged and made up the background, and nothing like the spike at the beginning of each occurs in the second half.
So, I have a few questions. First, when alpha particles hit beryllium, is the radiation of the neutrons isotropic, or are neutrons emitted in some anisotropic way? Related to that, once the neutrons hit the hydrogens in the moderator a few times and thermalize, what does the distribution of thermal neutrons look like? Is it more or less isotropic with greatest density right next to the source and diminishing in inverse-square fashion after that, or does it look different? If not, is it possible that putting a layer or two of HDPE between the sources and the target is a good idea?