The Pea and the Beach-Ball

Nuclear physicists describe the propensity of a nuclide to absorb a neutron in terms of a very small area, called a “barn”. A barn is a trillionth of a trillionth of a square centimeter (10-24 cm2).

Now imagine if we could inflate the size of a couple of nuclides, brothers actually, by a factor of two-and-a-half trillion.

These two nuclides, despite being from exactly the same element, would look very very different in their “size” to a thermal neutron.

Image Removed

One would be really big, about the size of a 36″ beachball. We would accurately conclude that a neutron would have a very good chance of being absorbed by this big guy.

Image Removed.

The other would be really small, about the size of a pea.

Now you might not believe that two brothers could be so different! Furthermore, you might be tempted to think that the beach-ball nuclide is the bigger one, and the pea-sized nuclide is the smaller one.

But you’d be wrong, because in the strange world of nuclear physics, the little one is big and the big one is little.

The nuclides I’m talking about are the two natural nuclides of lithium.

The little one, lithium-6, has three protons and three neutrons. It’s the beach ball.

The big one, lithium-7, has three protons and four neutrons. It’s the pea.

It’s the way they behave towards neutrons that’s the subject of this post. You see, they’re as different as Jekyll and Hyde. One wants the neutron, and the other one pretty much ignores it.

For the purposes of making energy from thorium, lithium turns out to be an very important element. That’s because lithium and beryllium fluorides make the finest mixture of salts into which we put our nuclear fuel as we operate the liquid-fluoride thorium reactor.

Beryllium only has one natural isotope. It doesn’t pay much attention to neutrons.

Fluorine also only has one natural isotope, which also doesn’t pay much attention to neutrons.

But lithium has two, and like the beach-ball and the pea, they look very different to the neutron. So it’s very important that we only use “peas” in the reactor instead of “beach-balls”. Fortunately for us, lithium is already mostly “peas”. Unfortunately for us, there’s still far too many “beach-balls”.

Natural lithium is about 92.5% lithium-7. But there’s enough lithium-6 in there that the mixture looks about like a “soccer ball” to a neutron. That’s much too absorptive.

If we were to enrich the lithium so that it was 99% lithium-7, things would be better, but still far too absorptive. Kind of like an apple.

At 99.9% we’d be down to the size of a strawberry, at 99.99% it would be about the size of a grape, and at 99.999% it wouldn’t be much larger than the pea.

But to get to that level of separation we’d have to do a lot of work. Separative work, to be more exact. We’d have to go from a lithium mixture that had 76,000 Li-6 per million down to a mixture that had only 10 per million. That’s a lot of separative work.

Leaving too much lithium-6 in the mixture could be a really bad idea too. Lithium-6 absorbs neutrons and turns into radioactive tritium gas, which is hard to capture and contain and if released into the environment will behave just like normal hydrogen. It becomes part of water (tritiated water) and depending on how much you ingest, it can give you an elevated dose of radiation.

In addition, it’s just plain wasteful of neutrons to leave much lithium-6 in the mixture.

So one of our key challenges in preparing the materials to make energy from thorium is learning how to separate lithium-6 from lithium-7. And despite what I’ve told you here, to a chemical separation process they don’t look like a pea and a beachball.

They look just the same.

20 thoughts on “The Pea and the Beach-Ball

  1. One thing I've never gotten that's still largely unexplained here: why can't we just irradiate the lithium with neutrons to turn the 6-Li into 7-Li? If it's so big and easy to target selectively with neutrons, then maybe that's the best way to separate it?

  2. I'm with Adam. It seems to me that you could do the math and figure out how long it would be until you enriched all the lithium 6 to lithium 7. Indeed, why not use an accelerator and spall neutrons, shoot them at a target batch of molten lithium until you had the appropriate proportion of Li7, then convert to the fluoride and use in LFTR. If you used a subcritical reactor, already equipped with a spalling apparatus, the 6-to-7 issue would go into your calculations.

  3. Go ahead and do the math and I think you'll see very quickly that neutrons are probably the most expensive way to "enrich" lithium. Along with all the tritium trouble that you'll make.

  4. Question answered. Thanks! I guess my idea will be good when we have a fusion reactor onsite to produce the neutrons and burn the tritium.

  5. What Kirk says is true. Neutrons are expensive. Note also that you wouldn't actually convert Li-6 to Li-7, you would burn out the Li-6, which of course would have the same effect as enriching uranium. When Li-6 captures a neutron it makes He-3 and H-3. This is the process in which tritium is produced. The thermal Li-6 radiative capture cross section (n,gamma) is small (40 millibarns), where as the (n,alpha) cross section is very large (900 barns). Interestingly, the radiative capture cross section of Li-7 is approximately the same as Li-6, but it does not have the issue of the (n,alpha) cross section.

    Note also, that in the operation of the molten salt reactor that the Li-6 will be burned out over time. However, Li-6 is also produced from fast neutron reactions with beryllium contained in the salt, so that it will reach an equilibrium value.

    Clever ploy putting the cute kids in the picture.

  6. I asked my biochemist friend if he had any ideas for this, and his two-minute idea was to do a lithium salt ion-spray in a strong electric field, the spray should split into two streams.

    Has anybody explored this yet?

  7. What is the best way to get lithium in the vapor state?
    Once in a vapor state, a tuned laser can selectively ionize the Li6 and it can be collected electrostatically.

    This is the approach that GE is using in the SILEX process that it is commerciallizing. It is also the approach that USEC used with the AVLIS process. (The physics of AVLIS worked fine but the U-Fe alloy used was very corrosive in the liquid state and caused maintenance problems)


  8. Kirk,

    I've just put the bad-news/bad-news entry below (for the UK that is) on my Blog: LFTRs to Power the Planet ( http://lftrsuk.blogspot.com/ ). Any chance of putting my Blog on your Blogroll?


    Email to the Feedback Page of 'physics world', the monthly journal of the Institute of Physics.

    I've just sent an email mentioning the October 2010 'Nuclear power The road ahead' issue, with 28 pages of erudite commentary on all forms of nuclear power and 100 words about LFTRs.

    Also mentioned is the potential death-knell for thorium in the UK, encompassed in Summary of The National Nuclear Laboratory's Position Paper: The Thorium Fuel Cycle. This – from the body advising our Government – is all our elected Representatives need, to shrug their shoulders and carry on with business as usual.

    See the details under the 'Sciences Apathy' tab.

  9. Why is it that every time I hear someone talking about how lithium-7 has a negligible neutron cross section, there's this little voice in my head that keeps saying "Castle Bravo?"

    I think Edward Teller said something to the effect of: "We made something of a mistake on that one."

  10. I'm taking one more shot at this. With the large %-difference in atomic weight (6 is 14% less than 7), it seems to me that if vaporized Li ions were accelerated in a field (just as fragments are separated in a mass spectrophotometer), that separation could be accomplished reasonably well.

    No matter how you do it it will be expensive. However, my understanding at this time is the molten salt carrier, once the offending isotopes are removed, is stable for the life of the system.

  11. So, what we need is a cheap source of neutrons. Neutrons are produced by the sun, but many of them tend to be absorbed before they reach the surface of the earth. I wonder if we could build a very large capillary neutron lens in space, focusing these neutrons onto a suitable target for enrichment (Lithium-6 for instance). The problem is, this is still expensive. Perhaps the lithium-6 and neutron lens could be placed over a terrestrial neutron source, such as under a place which receives many electrical storms (such as near the equator) or perhaps a nuclear waste dump. Note that all of these may be useful for enriching thorium. But I haven't worked out the number and energy of neutrons necessary for this lithium project to be feasible. I leave that to the skilled nuclear scientists among us.

  12. I believe this is a mostly solved problem.

    Thermonuclear weapons use the deuterium-tritium fusion reaction in the second stage; but bulk amounts of tritium is expensive, not very easy to store, not as dense as you'd like and has a fairly short half-life.

    Weapon designers overcame this by using lithium-deuteride as the fusion fuel. Tritium is made on-the-fly from lithium and neutrons; with most of the neutrons comming from the fissioning of a "spark-plug" of U-235 in the second stage.

    These people will go to great extents to reduce the size and weight of the weapons and they did this by enriching lithium-6; they don't want the peas, they want the beach balls.

    They apparently solved the problem of separating lithium-6 and lithium-7. Between 1954 and 1963 they created 442 metric tonnes of enriched lithium-6 at the Y-12 facility in Oak Ridge, with a peak rate of ~90 metric tonnes per year, using the "Column Exchange process"(COLEX).

    COLEX is an electrochemical process described as being similar to the way sodium and potassium metals and chlorine are produced in mercury cells. It exploits a slight isotopic selectivity in the exchange of lithium between a metallic mercury-lithium amalgam and lithium hydroxide.

  13. Lithium isotope separation by electrolysis http://library.sciencemadness.org/lanl1_a/lib-wwwhttp://library.sciencemadness.org/lanl1_a/lib-www
    "Taking account of the 7.3% abundance of Li6 and a assuming stripping by about 20% of the transport we estimate the rate of production of Li6 (95% pure) as roughly 30 grams per day for a unit with the power
    of about 300 x 450 x 1.3 = 1.75 x 10^5 watts. The cost of power only per gram of Li6 produced, at $O.005 per kwh, is roughly $O.70."

  14. Chuck, that's not going to happen. Lithium-6's high cross section for capturing neutrons and producing tritium is too useful for both good and bad things.

    Either lithium-6 'tails' will not be sufficiently enriched to be attractive for other uses, in which case you can't more than a percent or two of weight reduction in lithium batteries or they'll be stockpiled somewhere by government institusions.

    Lithium enriched in Lithium-6 is useful if a nation like India or Pakistan wishes to go from conventional nuclear weapons to thermonuclear weapons without having to develop their own lithium-6 enrichment facilities. The Castle Bravo test used 37-40% enrichment, it worked fine, with a significant amount of Li-7 also being converted(possibly the cross-section is higher for very energetic D-T fusion neutrons; I do not know).

    Given enough decades or centuries of hacking away at the problem there will almost certainly be some practical approach to fusion. Fusion reactors have a breeding problem of their own; they have to produce as much tritium as is lost in the reaction and due to radioactive decay. This is presumably easier with a lithium salt enriched in lithium-6; first generation fusion reactors may need to use lithium-6.

    D-D fusion has two branches, each with about 50% branch probability. D + D -> T + p+ and D + D -> He-3 + n. In tritium-lean fusion fuel/targets there would be enough additional tritium and neutrons from D-D fusion that it should be no problem using natural lithium in the tritium breeding blanket.

  15. Soylent,

    You raise an important point. The Li-enrichment processes were optimized to produce Li-6. How effective are they at enriching Li-7? When producing Li-7, it's OK to leave Li-7 in the tails to prevent misuse. The Li-6 can also be blended with natural Li.

  16. What is the cost of the salt in a Liquid Salt Reactor? The Lithium 7 will be by far the most costly raw material costing a mininimum of $700 per kilo just for the power (if the computations above are correct). How many tons of Lithium 5 will a Liquid Salt reactor need per kilowatt of power?

  17. I am new to this forum. Kirk's outstanding TechTalk presentation on nuclear waste given at Google has led me here.

    There are no answers to the Li-6 problem I could give, only more questions:

    1. Jess, I don't see what kind of fast neutron reaction would produce Li-6 from Be, could you please specify this? If there is constant Li-6 production and thus constant Li-6 burning, what would be the equilibrium of Li-6 (i.e. would it be reactor poisoning or would it be no problem for the neutron budget?) and (roughly) how much tritium would be produced (e.g. per kWh or GWa)?

    2. Is there any chance of realizing a LFTR concept that does not involve any enrichment issue and – more important for selling LFTR to the anti-nuke fraction – produces no tritium whatsoever?

Leave a Reply