Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Tritium self-heating

  1. Jul 25, 2010 #1
    Hi All

    Say I have a small ball of frozen tritium, about 2mm in radius. Does it heat up from its own decay since tritium decay releases ~325 W/kg? Or does its low-energy betas escape? How quickly (i.e. how far do they travel before being 'absorbed') do they thermalise? Is most of the decay energy in the electrons being spat out or the neutrino component?
  2. jcsd
  3. Jul 25, 2010 #2


    User Avatar
    Staff Emeritus
    Science Advisor

    A couple of beta energy spectrum are given here:

    The most probably energy of the beta particle is ~1/3 the maximum energy, but the decay of Bi-207 shows that the most probable energy is ~1/6 of maximum. It's been a whlle since I've looked into the details of beta decay, so possibly as the energy increases, there is a trend for the most probably energy to decrease as a proportion of maximum energy.

    The beta particles lose energy based on collisions with electrons or interactions with nuclei (brehmsstrahlung), which is a function of energy. One has to know the linear energy transfer (LET) rate. In addition, the beta emissions are dispersed throughout a solid, such that only a fraction of the energy is lost before the beta escapes. LET is a function of the density of a material and more so the electron density of the material, e.g., tungsten or lead are much better shielding material (absorbers) than copper, which is much better than aluminum. On the other hand, the tritium beta particle has low energy. However, I would expect most of the beta energy would leave a tritium sample if the size of the sample is on the order of 1 mm. Tritium is more effective if combined as metal hydride or bound in some organic compound.

    Crystalline solid hydrogen at 0.088 gm/cm3 is the lightest of all crystalline substances. (I believe that is under normal 1 atm pressure).
    Ref: http://hyperphysics.phy-astr.gsu.edu/hbase/pertab/h.html

    This group has measured beta decay of tritium -
    Fackler O, Jeziorski B, Kolos W, Monkhorst HJ, Szalewicz K.
    Accurate theoretical beta -decay energy spectrum of the tritium molecule and its neutrino mass dependence.
    Phys Rev Lett. 1985 Sep 23;55(13):1388-1391.

    More general information of beta (and alpha) decay
    http://oak.cats.ohiou.edu/~piccard/radnotes/alphabeta.html [Broken]

    Note that tritium decays to He-3. He-3 with Z=2 will attract an electron from a nearby hydrogen (tritium) molecule, and ulimately the beta source will attract electrons from nearby matter in which the beta particles are slowing down. Nature has a preference for charge neutrality.
    Last edited by a moderator: May 4, 2017
  4. Jul 26, 2010 #3

    Thanks for the interesting links & data. Tritium's density is somewhat higher than protium when frozen, at ~320 kg/m3. I am interested in knowing how a deuterium/tritium core about 2 mm in radius would heat a larger D/He3 mix around it, about 20 mm in radius, all wrapped in a thin layer of aluminium (small thickness.) The storage temperature should be ~3 K so the vapour pressure of the He3 is about 0.813 bar, but it's mostly in liquid form, inside a frozen deuterium 'honeycomb'. Stoichometric mix of all the components as it's an ICF fuel pellet. From what I can tell via a BOTE calculation is that the equilibrium temperature would be much too high for it to be indefinitely frozen without significant active cooling. Even then I am uncertain if the He3 would remain liquid, since I don't imagine the thermal conductivity of the deuterium/He3 mix is sufficient to get rid of the excess heat quick enough.
  5. Jul 26, 2010 #4


    User Avatar
    Staff Emeritus
    Science Advisor

    He-3 is never solid, and in D, it would be atoms or bubbles in a D-matrix. Perhaps the best way to develop it would be a DT mix, but then one has to wait until T decays to He-3.

    Otherwise, it's probably best to simply use He-3 gas and a neutral beam injector to feed a D-based plasma.
  6. Jul 26, 2010 #5
    Since I'm talking ICF then there needs to be a semi-solid target to undergo ablative compression. Liquid He-3 isn't excessively onerous to maintain, but the fuel pellet design I'm studying the characteristics of has the deuterium-tritium trigger, thus my inquiry.

    Alternatively, because He-3 can form Cooper pairs, then it might be possible to cause it to undergo collapse like the recently observed Ultradense Deuterium. In the ultradense state, the need for a tritium trigger is obviated.

    I found a mention of tritium self-heating in an experimental physics paper which involved making pellets of tritium. The tritium took a few minutes to melt, which isn't the kind of timeframe I wanted. Oh well.
  7. Jul 28, 2010 #6
    qrall, yes, tritium is self heating and the beta will not escape as long as the tritium is contained in any solid container. Beta particles will not even penetrate a peice of paper. For instance, a steel container about the size of a softball holding 10 grams of tritium at will actually feel quite warm to the touch. Tritium and DT mixes can be frozen and injected into plasmas as pellets. In order for this work, you must have a cryogenic system to continuously remove the heat and keep the DT mix at less than 20K so it will remain a solid. This is what we plan to do at ITER using a screw extruder pellet injection system.
    Last edited: Jul 28, 2010
  8. Jul 29, 2010 #7
    The ß- energy from H-3 decay is 18.51 keV. Using the Range and Stopping Power module in Nucleonica (www.nucleonica.net, free access but registration required) with the target solid hydrogen (density 0.08 g/cm2), the range is approximately 40 µm. So all the betas will be absorbed.
    Last edited: Jul 29, 2010
  9. Jul 29, 2010 #8
    Hey thanks for the software link. Much appreciated, but two corrections. First, the average energy of the betas is 5.7 keV, as the rest goes to the neutrino to make up the 18.51 keV you quote. Second, the mix is D-T which has a density of ~0.27 g/cm3. I'll plug in the figures and see how accurate it is.
  10. Jul 29, 2010 #9
    "Doesn't accept energies less than 10 keV"... kind of useless in this context. Oh well, I'll have to trawl through the reference textbook and see what it says.
  11. Aug 9, 2010 #10
    ..but if the 18 keV electrons are absorbed within the 2 mm pellet so also will all lower energies.
  12. Aug 19, 2010 #11
    Good point. I read that reference I mentioned here...


    ...and it pretty much confirms what you said. For some reason I misread the range you gave. *sigh* Oh well. Now I know. Thanks for all the responses. The frozen tritium extruder sounds interesting - I read an OSTI design study of it which was very handy.
    Last edited by a moderator: Apr 25, 2017
  13. Feb 21, 2011 #12
    Does anyone have any figures for the density of liquid or solid tritium?
    How does this compare to Ultra-dense tritium?
    Ah, and are there any proper publications on this?
    Last edited: Feb 21, 2011
  14. Feb 21, 2011 #13
    There's an Oak Ridge National Laboratory paper on Tritium and Deuterium which has their various properties listed at low temperature, which a Google search should dredge up. Look for a paper about tritium ice extruder or something like that. There's a table in there.

    If the theoretical work on ultradense deuterium is correct, then there can be no ultradense tritium - deuterium is bosonic and thus able to pack much closer than either protium or tritium. Helium-3 can form Cooper Pairs and so it might be possible to condense those in a similar fashion to UDD, but no one has done the experiment yet AFAIK.
  15. Feb 22, 2011 #14
    I will look for it thanks, if you happen to have a link that would be appreciated.

    I followed the work on creating ultradense deuterium by forming a metallic deuterium lattice, the density is rather increadable.

    I could not work out how this could be achieved with tritium, but wondered if I was missing something as I have read various postings about UDT. Incidentally there is theoretical "metallic hydrogen/protium " which is ultra dense, remember hydrogen is actually a metal.

    What I was considering is that as tritium exists as a gas in HT DT or TT state, could it piggyback onto Deuterium as part of the process for forming a metallic deuterium lattice?
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook