Posted by Mark Halper


Looking for lithium. Chemist Stephen Boyd cautions that lithium-7, a key element in the MSR salt FLiBe,    can be difficult to obtain.

When it comes to fission reactor designs, there’s nothing quite as safe, efficient, meltdown-proof, waste-light and proliferation-resistant as a molten salt reactor (MSR), many nuclear experts believe.

But as the community of MSR developers perfects its designs, it will have to overcome a number of engineering and materials challenges. One of those, according to a leading expert: Obtaining the elements to make the elixirs – the molten salts – that define the reactor.

First, a quick review. As it says on the label, a molten salt reactor (MSR) uses a liquid salt as the fluid that both carries the fuel (uranium, thorium or even a mix of plutonium among other possibilities) and serves as the coolant that picks up heat from a reaction and transfers it to a turbine.

It is the liquid nature that underlies all the reactor’s advantages, not the least of which is safety.  As guest blogger John Laurie pointed out here recently , liquid reactors – MSRs – cannot encounter a meltdown accident because they are already “pre-melted”. If things overheat, MSR designs allow the liquid fuel to drain harmlessly away into a holding tank.

A good MSR salt can continue to flow as a liquid at temperatures far above the operating temperature of conventional solid-fuelled, water-cooled reactors. MSR developers envision temperatures in the 700 degree C and 800 degree C range, a level that improves generating efficiencies and that makes the MSR highly useful as an industrial heat source.


Not all salts are up to the task, however. The one that many MSR developers believe best suited for the job is called lithium fluoride beryllium fluoride, or FLiBe. (It’s so much associated with MSRs that it’s the namesake of one well known MSR company, Huntsville, Ala.-based Flibe Energy).

But there’s potentially one big problem with FLiBe: Obtaining the lithium isotope called lithium-7 that FLiBe requires.

Lithium-7 occurs in natural lithium, where it is the main isotope, cohabitating with a little bit of another isotope, lithium-6. Common lithium is 92.5 percent “7” and 7.5 percent “6.”

The two isotopes together form regular lithium found in everyday items like cellphone and computer batteries. But separating them is not straightforward. According to chemist Dr. Stephen Boyd, chief technology officer of Aufbau Laboratories, Blue Point, N.Y., only two countries do it – China and Russia – and they use a separation process that relies on mercury, a hazardous substance that requires great care. Mercury is infamous in history for causing neurological disorders among 19th-century hat makers – giving rise to the term “mad hatter disease.”

The U.S. government abandoned the mercury method a number of years ago at its former lithium separation facilities at Oak Ridge National Laboratory in Tennessee. It recently announced a $120 million mercury clean up program there. Some reports have suggested that the inexplicable loss of a large amount of mercury from Oak Ridge also played a role in the decision to halt mercury-based lithium separation.

ThermonuclearBikiniAtoll DOE

Another use for lithium. The U.S. used lithium-6 and hydrogen to fuel this thermonuclear detonation on   Bikini Atoll in the Pacific Ocean on March 1, 1954. The Department of Energy’s defense-linked Y-12 group monitors lithium production.

So why was the U.S government extracting lithium-7 in the first place? Because it uses lithium-7 as a neutralizing agent – a pH balancer – in the small reactors that power the Navy’s fleet of nuclear submarines and aircraft carriers (commercial nuclear operators use it in the same way). Lithium-6 cannot be used in this process because it would transmute into potentially dangerous tritium (more on that in a moment).

The U.S. lithium separation process has been closely controlled by the federal government over the years, and not just because it wants to assure a supply of lithium-7 for its naval vessels.

Another reason: The enrichment of lithium into lithium-7 by definition also yields lithium-6, a substance with nuclear weapons links. Lithium-6 can be used to make tritium, a hydrogen isotope that is a fuel in hydrogen bombs (which work on the principles of fusion power, releasing energy by fusing tritium with another hydrogen isotope, deuterium).


With such deadly and national security implications, the government controls lithium-7 and lithium-6 enrichment through its Y-12 program, an Oak Ridge-based operation that resides in the Department of Energy but which serves national security and defense purposes.  As the DOE group’s website notes:

“Y-12 helps ensure a safe and effective U.S. nuclear weapons deterrent. We also retrieve and store nuclear materials, fuel the nation’s naval reactors, and perform complementary work for other government and private-sector entities.”

What this all suggests is that there’s an opening for a new lithium-7 extraction process. However, any company attempting such a development will have to work under the watchful eye of DOE’s Y-12 group.

That’s what Dr. Boyd has in mind at Aufbau.   Boyd, who wrote a guest blog here earlier this month in which he cautioned about other materials challenges facing MSRs, says he is developing a non-mercury process.  He declines for now to reveal details of his technology, but says he has been in contact with Y-12.

The knock-on effect for anyone in the MSR business is that they might find supplies of lithium-7 to be tight, at least until new potential supplies such as Aufbau’s or others come around.


That might be one reason why several MSR developers about whom I’ve written recently are considering salts other than the lithium-7 reliant FLiBe. A number of salts have the similar “eutectic” properties of FLiBe that minimize the chance of them solidifying (just like you don’t want your salt to boil, you also don’t want it to turn into a solid).

Canada’s Terrestrial Energy, for instance, is considering using sodium-based salts. Thorium Tech Solution is contemplating FLiNak, which is a combination of sodium, potassium and lithium. (There are reasons other than the lithium component for choosing a different molten salt. TTS has expressed concerns with FLiBe’s beryllium, an element that might disagree with the plutonium that will form part of TTS’s liquid fuel).

Each of these has its trade-offs. They can be less expensive and easier to obtain, but some of them can damage MSR plumbing.

Boyd believes it makes the most sense to stick with FLiBe for commercial, scientific and national-security reasons.

“FLiBe is a very good eutectic for MSRs,” he notes. “It’s one of the best, but you can run into the lithium problem right away.”

He also notes that there’s a business case for sticking with lithium-based FLiBe, because there are valuable markets outside of nuclear reactors for lithium isotopes and related products. He claims that lithium-6 sells for $1 million per kilogram.

Stay tuned for more on Boyd’s lithium enrichment ideas.  I met with him and with other MSR and materials mavens at the Thorium Energy Alliance Conference in Chicago last week,  a gathering that was full of bright ideas on the development of alternative and safe nuclear power. Among them: How one U.S. politician thinks thorium reactors developers can find funding. Watch for my reports.

Photos: Stephen Boyd by Mark Halper at Argonne National Laboratory outside Chicago, June 1, 2013. Bikini Atoll thermonuclear detonation from the U.S. DOE.


  1. Keith Pickering says:

    In his new book, Thorium: Energy Cheaper Than Coal, Robert Hargreaves is dubious of lithium in MSRs for a different reason. Li-7 will transmute into Li-6 in the reactor from the reaction Li-7 + n = Li-6 + n + n. The Li-6 will then form tritium, which presents a radiation release risk. The tritium risk can be managed, but that involves yet-another fuel salt processing step, along with its associated expense.

    • stephen B., Ph.D. says:

      Hi, Keith –

      You’re correct (and so is Robert), but you’ve both been correct for quite a long time. All PWRs (like in the Navy) use 7LiOH•H2O to moderate their water and they deal with the 3H. As a solid aside, I have developed a prototype which can pull the 3H out without opening the “box”.

      • Barry Osterman-Burgess says:

        Dr. Boyd,

        Can you provide any (basic, NDA-safe) details as to how the 3H-solution prototype works?

        • stephen B., Ph.D. says:

          Hi, Barry –

          Unfortunately, I can’t, other than to say it’s not gas-based. I would very much like to begin implementing R&D on this exciting project!

    • Mark Pawelek says:

      Li-7 + n = Li-6 + n + n

      That’s called a (n, 2n) reaction. I’m told it only happens to any significant amount in fast reactors. People propose using lithium only in thermal reactors, not fast. Fast reactors will get sodium chloride. Na-23 and Cl-37. Sodium is already nearly all Na-23. Chlorine is a 3 : 1 mixture of Cl-35, and Cl-37. Chlorine is easy to separate into 35 and 37, by fractionally distilling hydrogen chloride; a process dating from Victorian times.

  2. GreenEntropy says:

    How does this compare to using crown ethers for isotope separation? More information and papers on the topic may be found at:

    • stephen B., Ph.D. says:

      Hi, GreenEntropy –

      Crown ethers do have advantages, such as the ones being explored by the Chinese right now. I would submit, however, that the techniques I propose address the math (the exponentials that come with each successive “9”) and scaling issues a bit better.

  3. Pliftr says:

    We should wish Stephen Boyd much success in developing his technology.

    FLiBe could be probably best candidate among salts for MSR if not for the tritium problem resulting of the lithium-6 transmuting from beryllium. Two other troubles with FliBE are beryllium toxicity and the BeF2 related viscosity which increases in excess at temperatures approaching the eutectic melting point. On the other hand I was glad to find recently FliBe is expanding by about 1% only when freezing which will not create plumbing damage problem.

    Lithium-7 non-mercury, high-purity enrichment technology would surely strenghten FLiBe position vs other salts. Best of luck for your success, Stephen.


    • stephen B., Ph.D. says:

      Hi, Maciek –

      As was pointed out at the TEAC-5 conference this past week in Chicago, hosted by John Kutsch, Dr. Vidal illustrated that some of the fears associated with Be are overwrought. Yes, of course it must be handled with care, but Dr. Vidal is an expert with 30 years under his belt working with Be. Note for example, that it’s BeO dust which is toxic, due to its friable nature, yet it’s water insoluble. BeF2, on the other hand, is very hygroscopic. Knowledge of simple properties like this can go a long way toward the safe and long-term handling of these materials.

  4. Howard Caplan says:

    Hey Steve: Like the Beatles sang: so, you wanna save the world…
    I read the material your dad sent our way, and although I can follow the logic, I am ignorant of the why’s, not the what’s, meaning I really don’t “get” why Li7 can do what it does, and potentially can produce endless energy for at least another couple of decades. But his question was more like: do you recognize this guy, meaning you. Of course, you were here a year ago last March, and I remember you and your sidekick, who is indeed a genius, as are you. I have referred, in many of my daily comments to the NY Times opinion writers, to nanotechnology as the science that will change many of the problems in today’s world into solutions that will benefit us immensely. Hope you get underway in your endeavors and put nano to work for humanity. Howie

    • stephen B., Ph.D. says:

      Hi, Howie –

      Thanks for the comments. The specific 6Li and 7Li isotopes have very distinct nuclear properties. Those properties, in turn, can be exploited. With the 7Li, the “target” that an incoming neutron “sees” is extremely small. So, when using the chemical properties of Li (like in the baseLiOH•H2O), you’re reasonably safe to use the 7Li and permit it to get close to a neutron source.

      This is starkly different from the 6Li isotope, where if you hit that nucleus with a neutron of the right energy, you get a radioactive tritium.

  5. Jaro says:

    Glad to see chemist Stephen Boyd caution that lithium-7, a key element in the MSR salt FLiBe, “can be difficult to obtain.”
    The problem of “Difficult to obtain” materials, commonly known as “unobtainium,” generally translates to high costs.
    For example, the recent transfer of a vat of FLiBe salt from ORNL to the Czech nuclear research institute Rez is indicative: A complex logistical & political/regulatory process of transfer of nuclear materials suggests that it would have cost more for Rez to produce Li7 by themselves — whether by mercury or crown ethers processes.
    What was the total cost of that transfer ? …several million dollars ?
    Alternatives costing a few thousand $ certainly look appealing in that sort of background.

    Aside from that issue, the history of FLiBe and MSRs at ORNL may well be tied to DoE’s requirement to supply the US weapons stockpile with large amounts of tritium.
    ORNL demonstrated with the MSRE project that reactors using FLiBe salt are excellent producers of tritium.
    Unfortunately for ORNL, several high-power tritium production reactors were built at SRS instead (Savannah River Site).
    These SRS reactors were of the more conventional type, with solid fuel and water moderator. They used rods loaded with lithium, from which tritium could be easily harvested — in contrast to MSRE, where tritium was all over the place in the system plumbing and heat exchangers, and the reactor had a number of other issues, mainly related to corrosion and graphite moderator degradation (expected at power densities higher than those demonstrated in the experiment).

    With the tritium supply issue resolved, there was no need to continue funding MSR development at ORNL, and the project shut down (the earlier Aircraft Nuclear Propulsion project, ANP, also shut down, just as the “Fireball” or ART reactor was nearing completion of construction…)

    It would be great to see a new, economical method developed for Li7 production.
    But if cheap alternatives can be adopted instead, by using modern structural materials for reactors, then why bother ?

    • stephen B., Ph.D. says:

      Hi, Jaro –

      Excellent points you raise. We would bother, however, because 6Li and 7Li both have mature markets for research as well as for commercial uses. Adopting alternatives is fine, but it doesn’t obviate the need to continue moving forward on both basic and commercializable science like lithium enrichment.

      As far as implementation of Li-enrichment schemes, it probably just made good cost-benefit sense for Dr. Uhlir to obtain the 7Li-enriched FLiBe from ORNL, rather than set up his own. After all, ORNL are (unfortunately) not doing anything with it, which is a sincere disappointment.

  6. Alan Brown says:

    Neophyte question:

    There’s a lot of discussion over why 6Li is undesirable in terms of transmutation to tritium, but what are the actual consequences in a molten salt reactor if this is allowed to occur? (short term and long term)

    What could be done to mitigate the issue, other than isotopic seperation?

    Can issues be designed around?

    • Dan MacPhail says:

      I’m also a layman, and from the reading I’ve done I’m interested in whether there is any fundamental reason why unseparated lithium cannot be used as a corrosion inhibitor in the salt; if the only reason is to limit tritium production, would the commercial value of the tritium produced and the cost saving from not having to do isotopic separation not outweigh the extra cost of containing and sequestering the larger volume of tritium?

      • Stephen Boyd says:

        Hi, Dan –

        Lithium-6 has a much MUCH larger neutronic cross section than lithium-7. This means that it will transmute into tritium (an undesirable product) much more frequently than 7Li.

        You’re correct, by the way: tritium is of enormous value, but it’s extremely difficult to handle. Too much of it in the MSR will tend to have it migrate into the walls of the conduit pipe itself. There are much more efficient ways to either capture or produce tritium than errantly producing it in an MSR.

        • rick steffen says:

          HAS anyone explored GRAPHINE OR gold for the plumbing etc.???

          • Ivaylo Ivanov says:

            I was googling the same and could not find any research for gold as corrosion resistant metal for plumbing. I would guess that graphene will act similarly to graphite and will slow the neutrons to thermal spectrum and continue the nuclear reaction in the pipes.

    • Stephen Boyd says:

      Hi, Alan –

      Great question. Unfortunately, tritium behaves essentially identically to hydrogen, which is difficult to handle. The added difficulty is that tritium is radioactive. The nice thing about lithium enrichment is that the isotopes already have mature markets. Further, several researchers and I have come up with several additional (non-proliferative) uses for industrial-scale volumes of the two isotopes.

  7. gradin says: and the tritium byproduct could be extracted and sold to offset costs

Leave a Reply

Sign up for our Weinberg Next Nuclear Newsletter
* = required field

I am pleased to support the Alvin Weinberg Foundation’s mission to communicate honestly with the public and to raise awareness of the potential of this maligned energy source amongst campaigners and the media.

— George Monbiot


Our latest blog on the nuclear report from the Science and Technology Committee of the House of Lords. We need...
- Wednesday May 3 - 2:36pm

Recent Posts

Three Mile Island – the real disaster

by Suzanna Hinson (June 2nd, 2017)

Nuclear in Africa

by Suzanna Hinson (May 16th, 2017)

Engineers echo politicians: SMRs could help the UK post-Brexit

by Suzanna Hinson (May 11th, 2017)

Breaking the cycle of indecision: nuclear report by the House of Lords

by Suzanna Hinson (May 3rd, 2017)

Posts Archive


  • Economics (89)
  • Efficiency (54)
  • Policy (17)
  • Proliferation (32)
  • Regulation (8)
  • Safety (63)
  • Security (18)
  • Technology advances (23)
  • Uncategorized (53)
  • Waste (52)
  • © The Alvin Weinberg Foundation 2014
    The Alvin Weinberg Foundation is a registered UK charity. Charity number: 1155255
    The Alvin Weinberg Foundation web site uses cookies to record visitor patterns.
    Read our data protection policy

    Design by Tauri-tec Ltd and the Alvin Weinberg Foundation