This episode is brought to you by Brilliant. Contrary to popular belief, nuclear energy doesn’t give you superpowers, which is a bit disappointing. As a consolation prize though it does offer a near unlimited supply of clean and safe energy. So today we’re back to discussing the future of nuclear energy to focus in on Thorium, an energy source often heralded as the cleaner and safer successor to traditional nuclear power and the gateway to humanity’s independence from fuels that are short in supply or environmentally dangerous.
Many others view Thorium as little more than a pipedream, or even a dangerous distraction from pursuing other options like solar or wind or geothermal. Some argue it even distracts from improvements to more traditional nuclear fission or its long awaited upgrade, nuclear fusion. So which of these is right? Is it a Miracle power source, an over-hyped alternative, or a dangerous distraction? We will be exploring the science of thorium power and its future today, here on Science & Futurism with Isaac Arthur, and I am your host, the aforementioned Isaac Arthur.
Today’s gonna be a lengthy and maybe a bit of a technical discussion, so you may want to get a drink and snack, and don’t forget to hit the like and subscribe buttons while you’re at it. So we should probably start with the basics first, like what is Thorium, and what brought it into the public eye? Well Thorium, named for Thor, the God of Thunder and Lightning, is probably a good name for a substance that offers vast amounts of energy, and maybe also new atomic weapons if we’re not careful. And while many things helped popularize discussion of it as a future power source, sadly one of those was an old meme about a thorium powered car. I suspect that’s also what got this power source some of the cynicism often directed at it being over-hyped.
It is actually possible to make a nuclear-powered car, in a variety of ways, but none really stand out as economically sound or physically safe options. Of those that are it would probably be a RTG running on passive nuclear decay and charging the car’s batteries continuously, or some fission engine that did the same, could be throttled, and probably also ran on some super-rare transuranic isotope like Americium-242, which could permit a 5 kilogram reactor running on 700 grams of Americium. Such elements don’t naturally occur as they have a half-life vastly shorter than the age of Earth; in the case of Americium-242, just 432 years. Nonetheless, such elements and reactors could see a role even in personal vehicles in a century or two if the civilization is a post-scarcity civilization, with energy abundance and medical nanotechnology removing any health concerns about radiation, and on this show this is something we would expect to be true for the 22nd century, so a nuclear car running on rare isotopes is not off the table. It just won’t be getting served for dinner any time soon.
See these isotopes aren’t so much rare as non-existent. We have to make these isotopes, and typically with a breeder reactor. Which is also part of the process for using Thorium as a power source, but unlike Plutonium and the other Transuranic elements, thorium isotopes typically have half-lives similar to or longer than uranium. While its shortest lived isotope, Thorium-227, has a half-life of only 19 days, its longest lived isotope, Thorium-232, has a half-life of 14 billion years, the approximate age of the Universe, and about 3 times longer than Uranium’s longest-lived isotope, Uranium-238, which comes in at 4.5 billion years and is roughly the age of the Earth. This is the primary reason why Thorium is three times more abundant than uranium on Earth, indeed roughly as common lead, and almost all Thorium is Thorium-232, the element that interests us, whereas uranium is virtually all U-238, with a modest amount of the shorter lived U-235, which is fissile and has a half life of 700 million years, and also little bit of Uranium-234, about 50 parts per million, but with a half life of only a quarter of a million years.
Why there is any U-234 naturally occurring matters a lot for why Thorium should be useful to us. Now you have probably heard that Earth was created by a Supernova causing our solar system to form and most of our heavy elements are thought to have come from that Supernova. But not all would have, some coming from older ones or even from supernova remnants we have since passed through, and furthermore we cannot assume all supernovae produce equal amounts of each isotope of a given element.
Nonetheless we can say there were a great deal more equal amounts of the various isotopes of Uranium when they were made, and the date for that is usually put at about 6 billion years ago, well before Earth itself formed. So Thorium’s longest lived Isotope after 232, Thorium-230, only lives about 75000 years, which would be 80,000 half-lives in a 6 billion year period, If you cut something in half 80,000 times you end up with a number so small that even if every atom in the Universe had been composed of that substance initially it would be astronomically improbable a single atom of it was still around, anywhere in the entire Universe. Alternatively Thorium-232, with a half-life 14 billion years, would only have had about a quarter of its atoms decay in that time. Which gets us to why there is any Uranium-234 at all, since it’s only about three times longer-lived than Thorium-230, and indeed that is what Uranium-234 decays into. Both have such a short half-life, in geological and astronomical timelines, that even though Uranium-234 decays into Thorium-230, so many half-lives have passed that there shouldn’t be a single iota of either anywhere on this planet, and yet both are around in non-minuscule quantities.
This is because Uranium-238, with its half-life equal to Earth’s age, turns into Uranium-234, then Thorium-230, then on to Radium and Radon, then to Polonium, and finally on to Lead. In the grand scheme of things all those in-between decays represent only a tiny fraction of time compared to that spent first as Uranium-238, or afterwards as Lead, but it’s enough that we actually have some Uranium-234 and Thorium-230 on the planet, even though none of either would have still been around when the Earth formed, having decayed while making the journey from the Supernova or neutron star collision that formed them. Now something’s half-life is pretty indicative of how dangerous it is in terms of radiation, and so you need not concern yourself that there’s tons of Uranium and Thorium in your backyard. The stuff is chemically toxic if swallowed in large enough quantities, and its powdered metallic form is pyrophoric, meaning it ignites spontaneously in air.
But you’ve got about 40 micrograms of Thorium in your body, and eat about 3 micrograms a day. You have also got about 20 micrograms of Uranium in you too if you’re curious. Anyway, beyond that Thorium is pretty safe stuff to handle in its natural form, but it is still radioactive, even the ultra-long-lived Thorium 232, so while I agree with folks who say Thorium has advantages over Uranium as being a safer substance in terms of both handling and weaponization concerns, it’s still not something to casually play around with either. The big concern, and potential, comes from fission, and specifically from breeder reactions.
Same as an unstable atom can decay into a different atom spontaneously, we can slap it with something, typically a neutron, and have it break apart, and some of these substances will break apart in a way that emits more of whatever broke it in the first place, typically meaning it spat out a few neutrons after being hit by one. This can be used to send those neutrons to repeat the process on neighboring atoms and cause a chain reaction, and substances that are capable of this are known as fissile. Uranium-235 is Fissile; so is Plutonium-239, but it only has a half-life of 25,000 years so it isn’t naturally occurring either.
Plutonium-241 is also fissile and has a half-life of just 14 years. Uranium-233 is also fissile, and does better than plutonium at a half-life of 160,000 years, but that’s still not long enough to be naturally occurring. So we’ve got to make them.
This is why Uranium-235 is so useful for reactors and bombs, it is fissile and it is decently common, 0.72% of natural uranium. And since U-235 is slightly less dense than U-238, we can run it through a centrifuge to separate the two or at least produce a more concentrated selection of it; an enriched one. Though this is usually still mostly made of U-238. This results in the remaining part being somewhat depleted of U-235, usually half or more having been removed. Neither natural or depleted Uranium is particularly dangerous to handle, much the same as Thorium. Here’s the key bit though, while a material is fissile if it can absorb a neutron and break up to spit out more neutrons, causing a chain reaction, a lot of materials will just absorb a neutron and do nothing but sit there’s as a slightly different isotope of themselves, often a stable one too.
One of the things that makes lead handy as a radiation shield is that its three major isotopes, Lead 206, 207, and 208, can absorb a neutron and just become lead 207, 208, and 209 respectively, though Lead-209 will beta decay into ultra-long-lived Bismuth shortly thereafter. Beta decays are low-powered and basically harmless. Anyway, we don’t want these neutron absorbers in our reactor, or bomb, because they poison the fission process by sucking up some neutrons without spitting others out. However some substances do spit stuff out later.
For light elements we generally see an equal number of protons and neutrons, or close to it, but the bigger on the scale you go, the higher the ratio of neutrons to protons, and this permits very large numbers of variously stable isotopes. One example of which is Uranium-239, what you get when one of the many remaining U-238 atoms in an enriched Uranium fuel rod absorbs a neutron. Now U-239 is not a common word, mostly because it’s very short lived, and after 23 minutes turns into Neptunium-239 for a couple days, and then into Plutonium-239, one of our other fissile isotopes. Incidentally there are Uranium isotopes all the way from Uranium-214 to 242, but none of the first dozen last longer than a second. Plutonium-241 is also fissile but is made by hitting Plutonium-239 with a neutron to become 240, then doing it again to become 241, so it’s not a great way to make fuel, though Plutonium 241 has a better neutron absorption cross section than 239. Absorption cross section is the likelihood a passing neutron will be absorbed, more is better for fission.
As a side note, in general, isotopes with an odd number of neutrons are both more likely to absorb neutrons and undergo fission when they do. Thorium, Uranium, Plutonium, and Curium have 90, 92, 94, and 96 protons respectively, each an even number, so their isotope numbers indicate an odd number of neutrons when they’re odd, like Thorium-231, Uranium-233 and 235, Plutonium 239 and 241, Curium 243 and 245, each of which represent an odd number of neutrons and each of which is fissile and will have a lower bare critical mass for use as a bomb. Alternatively Americium-242 is the isotope of Americium, 95 protons, with an odd number of neutrons and its critical mass is put at around 9-18 kilograms, versus 50-155 kilograms for Americium-243. Folks more familiar with Thorium, who are probably getting impatient with me walking through this process, might be going “hey, I thought Thorium didn’t have any fissile isotopes” while others might be saying “See you can make a bomb out of it.” A thing to keep in mind for instance is that Thorium-231 has a half life of a day, and is principally made by the decay of Uranium-235, which is vastly more abundant and vastly more fissile. Lots of isotopes are technically fissile, but most are barely so, and often very short-lived.
That’s where Uranium-233 energy comes in, better known incorrectly as Thorium energy, which I say because we don’t sit around referring to bombs or reactors running on Plutonium-239 as being Uranium bombs or reactors. Thorium-232 is hyper abundant, being virtually all naturally available thorium atoms, and can absorb neutrons, but it doesn’t undergo fission. Instead it turns into Thorium-233, which still doesn’t undergo fission. It could absorb another neutron and become Thorium-234 in theory, but that has a short half-life of 22 minutes, after which it will decay into Protactinium-233, which has a 27 day half-life, then becomes Uranium-233, which is fissile. Technically it can decay into Thorium-229 at this point, continuing the chain, but Uranium-233 has a half-life of 160,000 years.
So our Thorium-232, which is busy absorbing neutrons, will have that atom spend a bit of time in those three mid-points before settling into Uranium-233. This is the Thorium Fuel Cycle, except we haven’t produced much power yet. Uranium-233 absorbs a neutron and sometimes turns into Uranium-234, about 6% of the time, which is good at capturing neutrons and will at some point do so and become U-235, which is of course quite fissile. Most of the time though, in 94% of interactions, the U-233 will fission from the neutron hitting it and break up into two smaller atoms, such as Xenon-137 and Strontium-94, plus 3 more neutrons, permitting the chain reaction as one of those neutrons might hit a neighboring U-233 to spit 3 more neutrons out, while another might be absorbed by some neighboring Thorium-232, starting the Thorium Cycle off again.
Incidentally, fission byproducts aren’t as uniform as decay products are. Most isotopes with a half-life decay in one specific fashion or perhaps two with a clear set percentage; fission is messier and can depend a lot on the speed or energy of the neutron hitting it. It’s also not perfectly documented, especially for these irregular processes like Uranium-233, since we basically have to pull all the materials out of the reactor and carefully scrutinize it for all isotopes, often in concentrations way less than one part per billion, and often track it backward by seeing some daughter isotope that’s a decay product of a fission byproduct. Much hunting for needles in haystacks, except atomic nuclei are very tiny needles.
One other note on this, we also have what we call parasites, sort of the opposite of fertile materials like Thorium-232, Uranium-234, and Uranium-238 which can absorb neutrons and end up as something fissile. During the production of Uranium-233 you inevitably get some Uranium-232 instead. This mostly happens when the Protactinium phase absorbs a neutron but can actually happen through multiple channels, both during the normal Thorium Fuel Cycle or even off the tiny amount of Thorium-230 rather than 232 in any given Thorium sample.
U-232 has a good neutron absorption cross section for fission, which you would think would be good for a fission reactor, but mostly it interferes with the Thorium Fuel Cycle and also is highly radioactive, with a mere 70 year half-life and being a powerful gamma ray emitter. This is one of the big problems with Thorium as a power source, once it’s irradiated – which is to say, once we start blasting it with neutrons: that U-232 starts emerging and represents a significant handling and safety burden. Another problem is that it isn’t really ideal for a fast reactor, and of course requires us to explain what a fast reactor is. Most neutrons coming out of these reactions are moving very, very fast, frequently a couple percent of light speed, or thousands of kilometers per second, one of the reasons we like it for spaceship drives, see our episode “The Nuclear Option”. We often slow these neutrons down to a thousandth that speed, something more like what a fighter jet might move at, via neutron moderation.
This is where we have a space between the fuel rods filled with something like water or graphite that neutrons can whack around and slow down by hitting the molecules, but generally don’t get absorbed by them. This heats that moderator up and often allows us to generate power by moving that moderator through a standard heat engine or for safety’s sake, using it to heat up something which heats up another fluid to avoid radiation getting out. That’s how what we call a “Thermal Neutron reactor” works, the neutrons are slowed to speeds of hot materials rather than fast relativistic ones.
A “fast reactor” uses those fast relativistic ones, being short for “Fast Neutron reactor.” But some materials can absorb even a high-speed neutron, called fast neutron absorption, and that’s ideal for the typical breeder reactor where you absorb neutrons to make a new substance rather than direct fission, which is what we do with Thorium and what we do with Uranium to make Plutonium, and the latter cycle is better at that. That’s a downside but not a fatal one, but reminds us that while U-235, the workhorse of fission reactors, is pretty rare compared to Thorium-232, U-238 is more than a hundred times more common than U-235, and like Thorium-232 has a breeder cycle, going to Plutonium-239, which cuts into the often proclaimed edge of Thorium as vastly more abundant. It is more abundant than Uranium-238 but not that much more so, but U-238 is for making Plutonium which is quite easy to weaponize, which Thorium is often proclaimed not to be.
Of course you can weaponize thorium power. U-233 can be made into a bomb for instance, which the U.S. experimented with once during the 1950’s, but it is arguably more proliferation resistant.
As I often say on the show though, power is power, and you can weaponize any power, even if just by generating it to run your bomb factories or your economy to afford more and better armed troops. The main path for weaponizing this is to be cycling the reactor fuel to remove the Protactinium from the Neutron Flux for a few months until it’s decayed into U-233, preventing most of the U-232 formation that makes U-233 non-ideal as a nuclear bomb, and Protactinium can be chemically separated from the Uranium and Thorium. However, hijacking lots of small thorium reactors to get that fuel to separate the tiny amount of Protactinium out is a lot of effort to get a bomb. And while I dislike folks calling Thorium safe from weaponization, I really haven’t heard of a good method for weaponizing it, albeit in all fairness that might be because it’s classified.
Even if the reactors were all over the place, that doesn’t seem any easier than sourcing material for a regular atomic bomb, though folks can be surprisingly inventive so we shouldn’t write it off. Especially given that one of the more popular thorium reactor designs relies on molten liquid salt in the core, not the typical image of hard metal rods or pellets, which is handy in large part because it allows us to chemically separate out fission byproducts. Incidentally it’s important to remember that while you can separate isotopes of the same element by centrifuging them, all isotopes of an element act the same chemically, so a process that dissolves Uranium works on U-235 and U-238 or U-233, same for separating out Protactinium from Uranium and Thorium, which would let you get purer Uranium-233. A pure Uranium-233 is much better for making bombs out of. For anyone who is curious, narrating a script full of Isotopes is a worse tongue twister than that time I helped Peter Piper pick a Peck of Pickled Peppers.
So yes, Thorium is somewhat proliferation-resistant, compared to U-235 or Plutonium, but I don’t like to tout Thorium for this reason because I think it is surrendering to the flawed premise that controlling one particular style of weapon or attack at the cost of a major economic or environmental advantage is a good deal. Not that I would encourage folks to get blasé about radiation or atomic weapons but if you are a channel regular then you are familiar with the wide array of doomsday devices potentially available to us in the not-too-distant future. So those are two often-cited advantages to Thorium I feel need caveats; the weapon proliferation edge isn’t maybe that advantageous, and Thorium’s abundance as a big edge only matters for classic U-235 reactors. Truth is, there’s more than enough U-238 to keep our civilization running on Uranium-Plutonium breeder reactors for centuries even without extracting the multi-million year supply of it in our oceans and seabeds, let alone extreme mining of our mantle or space, and Uranium is valuable enough in terms of energy density to make mining it from space profitable, especially with the sort of nuclear propulsion systems we looked at in our episode “The Nuclear Option.” Again, Thorium is more abundant in our crust but both are quite abundant enough, and uranium is much more abundant in our seawater which is currently a bit pricey to extract in terms of energy, though not too bad compared to normal mining.
But if you’re running an economy on Uranium and Thorium it’s a pretty minimal extra cost and would work well with both coastal plants’ cooling towers, and nuclear power for water desalination. However, realistically, if we haven’t worked out fusion in a few centuries when easily available Uranium might be running low, then buying ourselves three times the supply by switching to Thorium is not really a winning policy, in favor of something a bit more long term renewable like solar. Though again keep in mind we can access millions of years worth of either Thorium or Uranium, it's just the Sun lives for billions of years and gives off way more power. Two downsides of Thorium often cited that I’d dispute though are that it is tricky to handle, and we aren’t as experienced with it. Both are true, we do have way more experience with other types of reactors and uranium as a fuel of course. Needless to say that’s only a temporary downside and not a good reason to avoid it as a fuel, but it is to countries or companies that need make power now, and on a budget, so it’s valid but manageable by putting effort into it, same as we did for solar, which was a horrible power supply economically until this last decade, but is now the cheapest option in many parts of the world.
Also needless to say, with other power options besides traditional nuclear to compete with, taking a lot of time to learn to better handle and work with Thorium in terms of practical experience is a real downside. Thorium is tricky to handle too, if not ridiculously so. As I mentioned earlier, thorium can be pyrophoric, burning when exposed to air, and needs temperatures higher than uranium to melt and purify. But again this doesn’t really seem like a big limitation so much as a matter of developing the tools and infrastructure to do it.
A big part of the issue with time is the simple licensing process for an entirely new fuel source, which could take a decade or more. Years of modelling in computers has to be done, alongside physical tests, and then it takes years to review the license applications for different parts of the process. If one avoided any and all regulations, you could probably do it pretty quick, but that’s not exactly ideal for safety or the environment. On the other hand, some might argue that risk is small compared to current damage being done from other power sources.
Personally, the Thorium-Uranium power production process I like the most at the moment is with MSRs or Molten Salt Reactors. Sometimes called a Molten Salt Fueled Reactor to differentiate them from those using molten salts as the coolant or working fluid, here the fission fuel is in the molten salt. I mentioned moderators earlier to slow neutrons down for capture, and that’s not limited to water.
A big stream of molten matter cycling through your core is advantageous for many reasons, some of which we discussed in our recent episode on the Future of Fission, but one is that you can keep everything lower pressure and you can cycle it out so that you don’t need to load your reactor up with a ton of fissile material to keep the reactor operating a long time. That allows for much smaller reactors when it comes to shielding and containment; it’s not some giant pressure cooker full of tons of superhot radioactive material. This is where the LFTR or ‘Lifter’ Liquid Fluoride Thorium Reactor comes in, and it’s probably the most touted Thorium reactor concept. It’s got some variations too, and isn’t limited to Thorium designs. Fluoride melts at 135 Celsius, or 275 Fahrenheit, not that much above water’s boiling temperature, which is part of why it's appealing as the medium.
Also when Fluorine atoms absorb a neutron they just turn into Neon-20, which is stable and a Noble Gas so obviously very easy to remove. Fluorine also bonds to Uranium to create uranium tetrafluoride, sometimes called green salt, which becomes our core material doing the fission. Its melting point is 1036 Celsius and it is insoluble in water which means if your reactor gets mashed you’re not spraying the vicinity with superhot radioactive and toxic gas or something which can dissolve in rain or water to run all over the place. MSRs – thorium or uranium – have a lot of engineering upsides and downsides we will bypass today, but we’re discussing what we have thus far in terms of the mechanics to clarify why Thorium often gets discussed as usable in small reactors. If it helps conceptually it’s a lot like why your smart phone does not have a gas engine, even though that would be vastly better than a battery, and why even though electric cars are finally making it into market as real competitors, nobody is mass manufacturing battery run planes or freight trucks (or at least not yet). Thorium is not in any way more energy dense than Uranium; they’re about the same.
Most of that compactness is coming from the MSR design, not whether it runs on thorium-232 or uranium-238. Critically though, Thorium allows Thermal Breeding, or slow neutron breeding, in the context of those neutrons only moving at hot thermal speeds of atoms and not a fraction of light speed like fast neutrons do. Which means the amount of fissile material you have to put in to start the process up is lower, though Fast Breeders always make more fissile material as they have better neutron economy, which is a plus for Uranium-Plutonium over Thorium-Uranium.
And while these are small but not super small, there are other technologies which might allow even tinier ones. As an example in our Death Rays episode I mentioned how a neutron reflective substance would be great for making neutron beams, which was our lead candidate for a solid death ray, and that same sort of material or metamaterial, if we found one, would utterly alter the fission landscape and start letting us contemplate man-portable reactors, maybe even pocket-sized stuff, and there are other technologies that might permit similar advantages for thorium. So thorium holds a lot of promise and might have other chances to shine as other new technologies emerge. Though of course the reverse is true too; we’re only one miracle technology away from working fusion, or ultra-cheap space-based solar panels, or extremely high-performing batteries. Is there a future for Thorium? I think so, I think the path forward for humanity for this century at least is to rely on a mix of nuclear and solar as our two principal power sources – though not exclusive, just principal.
In that context I think with a bit of work Thorium MSRs like LFTR offer one of the better routes for rapidly mass producing and deploying these reactors in the longer term, maybe for replacing the reactors we already have with safer options. They may not be the miracle power source many claim, partially because we have so many other impressive power options getting explored and prototyped too, but they do offer a lot of advantages. In the end, an economically abundant and ecologically safe power source would be a miracle all by itself, and while we can debate if other power sources might be as good or better in that regard as Thorium, there is no doubt Thorium could fill that role.
So the future of thorium might be a bright one, and if it is, it will make for a bright future for humanity. If you’re looking to learn more about nuclear energy and nuclear reactors, a hands on approach is always best but with radioactive materials we need to be a bit more metaphorical about handson, and there’s some excellent lessons on those topics in Brilliant’s Course Physics of the Everyday. I want to stress that though because hands on learning and interactive learning, are simply the best approach to learning a topic, any topic, and Brilliant is a website and app built off the principle of active problem solving — because it takes more to learn something than just watching it — to really learn something, you have to do it. Over the last year, Brilliant has built a whole new platform for their courses that takes interactivity to the next level. To see examples of that, check out their courses on Pre-Algebra, Mathematical Fundamentals, Algorithm Fundamentals, and the newly released course on Scientific Thinking, with many more under construction.
They are simply the best online learning platform I’ve ever seen for Math, Science, and Computer Science, and there’s no secret to why: You learn best while doing and solving in real-time, not by long lectures or memorising formulas and facts, and Brilliant understands that and has something for everybody — whether you want to start at the basics of math, science, and computer science, or go straight to advanced material. If you'd like to join me and a community of 8 million learners and educators today, click the link in the episode description down below or visit: brilliant.org/IsaacArthur. So I recently had the chance to give a presentation on Human-Machine Teaming at MIT Lincoln Labs and that was a fun and exciting honor to speak to that group as well as to pick their brains for more information.
Preparing for that as well as the feedback from the Q&A afterward has given us some episode topics, and next week we’ll have the first of those episodes, Human-Machine Teaming, and the second, Criminal AI, will be out toward the end of October. In a somewhat similar vein, we often discuss the difficulty of hiding civilization or even spacecraft in regard to the Fermi Paradox, and in three weeks we’ll dive into that in detail in Fermi Paradox Hidden Civilizations but before that we’ll have our mid-month scifi Sunday episode on Stealth Spaceships on September 12. Then we’ll celebrate the 7th anniversary of our original episode on Megastructures with a look at Megastructure Death on September 16th. Then we’ll close the Month out by asking if it is possible for a civilization to exist without money and what that might look like.
Now if you want to make sure you get notified when those episodes come out, make sure subscribe to the channel, and if you enjoyed the episode, don’t forget to hit the like button and share it with other. If you’d like to help support future episodes, you can donate to us on Patreon, or our website, IsaacArthur.net, which are linked in the episode description below, along with all of our various social media forums where you can get updates and chat with others about the concepts in the episodes and many other futuristic ideas.
Until next time, thanks for watching, and have a great week!
2021-09-03