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

Radiation to electricity

  1. Mar 4, 2017 #1
    Is there any way to convert ionizing radiation to electric energy without converting it to heat in the meantime?
    I don´t mean avoiding losses as heat - these are unavoidable - but direct conversion of some energy to electricity, by a mechanism that does not depend on buildup of heat.
    For example, batteries convert chemical energy to electricity, and photoelements convert visible light to electricity. Anything for ionizing radiation?
     
  2. jcsd
  3. Mar 4, 2017 #2

    berkeman

    User Avatar

    Staff: Mentor

    https://en.wikipedia.org/wiki/Magnetohydrodynamic_generator

    :smile:
     
  4. Mar 4, 2017 #3

    Astronuc

    User Avatar
    Staff Emeritus
    Science Advisor

    Basically, one is describing charge separation, i.e., can one expel electrons from a mass (emitter) and collect the electrons on another mass (collector) at some constant rate (ideal) such that one maintains a constant voltage/current, which is the idea behind beta-voltaics.

    https://en.wikipedia.org/wiki/Betavoltaic_device
    http://large.stanford.edu/courses/2013/ph241/harrison2/
    https://www.technologyreview.com/s/416312/a-25-year-battery/
    https://www.energy.gov/sites/prod/files/2015/08/f26/Cabauy Tritium Focus Group Presentation.pdf

    berkeman cited MHD, in which the working fluid/gas in a strong magnetic field would be ionized by radiation resulting in charge separation, but recombination is a problem if the separation of emitter and collector are large compared to the range of the radiation.
     
  5. Mar 4, 2017 #4
    Is there any semiconducting device capable of converting kinetic energy of fission fragments into electricity?
     
  6. Mar 4, 2017 #5

    Astronuc

    User Avatar
    Staff Emeritus
    Science Advisor

    Fission fragments have very short range in solids. Fission spikes are generally less than about 7 microns, or effectively less than a typical nominal UO2 grain diameter. Converting fission fragments in a thermo-electric semiconductor would be problematic from that standpoint of accumulating a variety of chemical species in the semi-conducting material, in addition to the neutron activation and radiation effects due to neutron and gamma radiation. Most thermoelectric concepts I've seen involve liquid metals removing heat from the core and passing it though an ex-core heat exchanger. I've also seen in-core thermionic concepts, which have similar issues regarding activation/transmutation and radiation effects.
     
  7. Mar 5, 2017 #6
    How does this compare with alpha particle range? Alpha particles have lower mass, but also lower energy...
     
  8. Mar 5, 2017 #7

    anorlunda

    Staff: Mentor

    Did you ignore the beta voltaic device that @Astronuc linked in #3?

    BTW: @Astronuc 's very impressive knowledge of nuclear tech seems to be encyclopedic.
     
  9. Mar 5, 2017 #8
    Beta radiation is not kinetic energy of fission fragments.
     
  10. Mar 5, 2017 #9

    etudiant

    User Avatar
    Gold Member

    If memory serves, some of the more advanced fusion reactions yield entirely charged particles whose energy could in theory be harvested by suitable electrostatic fields. There were some studies of a possible demonstration using the fragments of collisions in an accelerator, but nothing was ever reduced to actual practice that I know of.
    Conceptually, this idea of direct electrical energy harvesting is a lot more appealing than splitting atoms to boil water, but there are technical obstacles.
     
  11. Mar 6, 2017 #10

    Astronuc

    User Avatar
    Staff Emeritus
    Science Advisor

    It's been a few decades since I've done calculations, but looking around I found some data at NIST.

    http://physics.nist.gov/PhysRefData/Star/Text/ASTAR.html (on can enter iron (Fe)), select No graph, and submit. One obtains a table of alpha particle range by energy.

    The range in g/cm2 can be divided by density in g/cm3 to get range in cm. It looks like the range of an alpha particle of energy 1 MeV is about 0.18 mm in iron/steel, and for a 4 MeV energy, the range is about 0.72 mm. I plotted range as a function of energy on a log-log plot, and there are two linear portions with a transition in between. So the range of alpha particles is considerable longer than fission products.

    I haven't verified the numbers, so consider the ranges preliminary.
     
  12. Mar 7, 2017 #11
    Note that a never run reactor would have a background of alpha decays.
     
  13. Mar 7, 2017 #12

    Astronuc

    User Avatar
    Staff Emeritus
    Science Advisor

    Yes, but the activity is very low. When fuel assemblies are manufactured, there is very little heat generated, and the fuel is stored without shielding in open air. There is no cooling necessary. The half-lives of U-235 and U-238 are very long, 704 million and 4.468 billion years, respectively. In processing uranium ore, the decay products are removed.
     
  14. Mar 7, 2017 #13
    ...except uranium 234, because it is uranium.
    In natural uranium, the composition is approximately:
    140 parts of U 238 to 1 part of U 235
    18 000 parts of U 238 to 1 part of U 234 (because U 234 is a daughter of U 238 and that´ts the ratio of half-lives)
    therefore 130 parts of U 235 to 1 part of U 234.
    But in terms of alpha decay rates, it is approximately:
    22 decays of U 238 to 1 decay of U 235
    1 decay of U 238 to 1 decay of U 234 (because U 234 is daughter)
    therefore 22 decays of U 234 to 1 decay of U 235.
    Now, while full uranium series is 8 alpha and 6 beta decays, U 234 is long-lived (240 000 years). Until such time as U 234 shall build up, the decay of U 238 to U 234 contains just 1 alpha and 2 beta decays.
    U 234 decay series is 7 alpha and 4 beta decays, but ionium is long lived (75 000 years). Until such time as Th 230 shall build up, the decay of U 234 to Th 230 contains just 1 alpha decay.
    Full actinium series contains 7 alpha and 4 beta decays, but no daughter is U and Pa 231 is long lived (32 000 years). Until such time as Pa 231 shall build up, the decay of U 235 to Pa 231 contains 1 alpha and 1 beta decay.

    What happens when U is enriched?
    When U 235 is separated from U 238, U 234 is even lighter than U 235!
    For example, if the ratio of U 235 to U 238 is increased 8 times, from 1:140 to 1:17,5, what happens to the ratio of U-234 to U-235?
    Does it remain the same, at 1:130? Or increase 2 times (cubic root of 8), becoming 1:65?
    Remember, the activity ratio of U 234 to U 235 begins at 22:1. Does enrichment cause further increase of U 234 activity relative to U 235?
     
  15. Mar 7, 2017 #14

    mheslep

    User Avatar
    Gold Member

  16. Mar 7, 2017 #15
    Interesting thread,
    I go with better technology for collecting solar power,
     
  17. Mar 7, 2017 #16

    Astronuc

    User Avatar
    Staff Emeritus
    Science Advisor

    Natural U is approximately 0.0054% U-234, 0.72% U-235 and the balance (~99.27%) U-238.

    I looked at some of my notes on enriched U for commercial fuel. I have assays giving U-234, U-235 and U-238. I did a quick fit of the data, which yielded U-234 (%) = 0.0006 * (U-235 (%))2 + 0.007 * (U-235 (%)), so the fraction of U-234 increases with a slightly parabolic trend with enrichment (% of U-235). At maximum commercial limit of 5% U-235, the maximum expected U-234 proportion would be ~0.05%, and U-238 is about 94.95%.

    The fraction of U-234 is not significant with respect to alpha activity. The half-life is ~245500 years, all the decay products have been chemically removed, and the decay of U-234 from mine to conversion and enrichment is not significant.
     
  18. Mar 8, 2017 #17
    To the contrary, it is highly significant. After all, even before enrichment, U-234 has as much activity in terms of alpha decay count as U-238 (although U-238 has 2 additional beta decays that U-234 does not have) and 22 times as much as U-235. Enrichment removes U-238, so the activity of U-234 becomes the most significant activity present.
     
  19. Mar 8, 2017 #18

    Astronuc

    User Avatar
    Staff Emeritus
    Science Advisor

    I've handled enriched pellets, and they are cold to the touch, as in room temperature. There is no significant thermal energy available. I've stood next to barrels containing 500 kg of enriched UO2, and they have no significant temperature.
     
    Last edited: Mar 8, 2017
  20. Mar 8, 2017 #19
    Sure.
    The squash court pile was something like 5,4 t metallic U - unenriched - and 45 t "uranium oxide". Which one? UO2 or U3O8?
     
  21. Mar 8, 2017 #20

    Astronuc

    User Avatar
    Staff Emeritus
    Science Advisor

    I would expect it was UO2, which has a higher density of uranium. I'll have to dig into available technical documents.

    From a brief description from U of Chicago: "Chicago Pile Number One, or CP-1 for short, consisted of 40,000 graphite blocks that enclosed 19,000 pieces of uranium metal and uranium oxide fuel." So uranium oxide is not very specific.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted