Cold Fusion is Back (there's just one problem)

[This is a transcript of the video embedded below. Some of the explanations may not make sense without the animations in the video.]



Cold fusion could save the world. It’d be a basically unlimited, clean, source of energy. It sounds great. There’s just one problem: it’s not working. Indeed, most physicists think it can’t work even in theory. And yet, the research is making a comeback. So, what’s going on? What do we know about cold fusion? Is it the real deal, or is it pseudoscience? What’s cold fusion to begin with? That’s what we’ll talk about today.

If you push two small atomic nuclei together, they will form a heavier one. This nuclear fusion releases an enormous amount of energy. There’s just one problem: Atomic nuclei all have a positive electric charge, so they repel each other. And they do so very strongly. The closer they are, the stronger the repulsion. It’s called the Coulomb barrier, and it prevents fusion until you get the nuclei so close together that the strong nuclear force takes over. Then the nuclei merge, and boom.

The sun does nuclear fusion with its enormous gravitational pressure. On earth, we can do it by heating a soup of nuclei to enormous temperatures, or by slamming the nuclei into each other with lasers. This is called “hot nuclear fusion”. And that indeed works. There’s just one problem: At least so, far hot fusion eats up more energy than it releases. We talked about the problems with hot nuclear fusion in this earlier video.

But nuclear fusion is possible at far lower energy, and then it’s called cold fusion. The reason this works is that atomic nuclei don’t normally float around alone but have electrons sitting in shells around the nucleus. These electrons shield the positive charges of the nuclei from each other and that makes it easier for the nuclei to approach each other.

There’s just one problem: If the atoms float around freely, the electron shells are really large compared to the size of the nucleus. If you bring these nuclei close together, then their electron shells will be much farther apart than the nuclei. So the electron shells don’t help with the fusion if the nuclei just float around.

One thing you can do is strip off the electrons and replace them with muons. Muons are basically heavier versions of electrons, and since they are heavier, their shells are closer to the nucleus. This shields the electric fields of the nuclei better from each other and makes nuclear fusion easier. It’s called “muon catalyzed fusion”.

Muon catalyzed fusion was theoretically predicted already in the 1940s and successfully done in experiments in the 1950s. It’s cold fusion that actually works. There’s just one problem: muons are unstable. They must be produced with particle accelerators and those take up a lot of energy. The muons then get mostly lost in the first fusion reaction so you can’t reuse them. There’s a lot more to say about muon catalyzed fusion, but we’ll save this for another time.

There’s another type of “cold fusion” that we know works, which is actually a method for neutron production. For this you send a beam of deuterium ions into a metal, for example titanium. Deuterium is a heavy isotope of hydrogen. Its nucleus is a proton with one neutron. At first, the beam just deposits a lot of deuterium in the metal. But when the metal is full of deuterium, some of those nuclei fuse. These devices can be pretty small. The piece of metal where the fusion happens may just be a few millimeters in size. Here is an example of such a device from Sandia Labs which they call the “neutristor”.

The major reason scientists do this is because the fusion releases neutrons, and they want the neutrons. It’s not just because lab life is lonely, and neutrons are better than no company. Neutrons can also be used for treating materials to make them more durable, or for making radioactive waste decay faster.

But the production of the neutrons is quite an amazing process. Because the beam of deuterium ions which you send into this metal typically has an energy of only 5-20 kilo electron Volt. But the neutrons you get out, have almost a thousand times more energy, in the range of a few Mega electron Volt. It’s often called “beam-target fusion” or “solid-state fusion”. It’s a type of cold fusion, and again we know it works.

There’s just one problem: The yield of this method is really, really low. It’s only about one in a million deuterium nuclei that fuse, and the total energy you get out is far less than what you put in with the beam. So, it’s a good method to produce neutrons, but it won’t save the world.

However, when physicists studied this process of neutron production, they made a surprising discovery. When you lower the energy of the incoming particles, the fusion rates are higher than theoretically expected. Why is that? The currently accepted explanation is that the lattice of the metal helps shielding the charges of the deuterium nuclei from each other. So, it lowers the Coulomb barrier, and that makes it more likely that the nuclei fuse when they’re inside the metal. This isn’t news, physicists have known about this since the 1980s.

But if putting the deuterium into metal reduces the Coulomb barrier, maybe we can find some material in which it’s lowered even further? Maybe we can lower it so far that we create energy with it? This idea had been brought up already in the 1920s by researchers in the US and Germany. And it’s what Pons and Fleischman claimed to have achieved in their experiment that made headlines in 1989.

Pons and Fleischman used a metal called palladium. The metal was inside a tank of heavy water, so that’s water where the normal hydrogen is replaced with deuterium. Ponds and Fleischman then applied a current going through the palladium and the heavy water. They claimed this created excess heat, so more than what you’d get from the current alone. They also said they’d seen some decay products of fusion reactions, notably neutrons and tritium. Everyone was very excited.

There was just one problem... Other laboratories were unable to reproduce the claims. It probably didn’t help that Pons and Fleischmann were both chemists, but nuclear fusion has traditionally been territory of physicists. And physicists largely think that chemical reactions simply cannot cause nuclear fusion because the typical energies that are involved in chemical processes are far too low.

A few groups said they’d seen something similar to Ponds and Fleischman, but the findings were inconsistent, and it remained unclear why it would sometimes work and sometimes not. By the early nineties, the Pons and Fleischmann claim was largely considered debunked. Soon enough, no scientist wanted to touch cold fusion because they were afraid it would damage their reputation. The philosopher Huw Price calls it the “reputation trap”. In fact, while I was working on this video, I’ve been warned that I, too, would be damaging my reputation.

Of course not everyone just stopped working on cold fusion. After all, it might save the world! Some carried on, and a few tried to capitalize on the hope.

One such case is that of Andrea Rossi who already in the 1970s said he knew how to build a cold fusion device. In 1998, the Italian government shut down his company on charges of tax fraud and dumping toxic waste into the environment. In the mid 1990s, Rossi moved to the USA and by 2011, he claimed to have a working cold fusion device that produced excess heat.

He tried to patent it, but the international patent office rejected the application arguing that the device goes “against the generally accepted laws of physics and established theories”. A rich Australian guy offered $1 million to Rossi if he could prove that the device produces net power. Rossi didn’t take up the offer and that’s the last we heard from him. There’s more than one problem with that.

In 2019, Google did a research project on cold fusion and they found that the observed fusion rate was 100 times higher than theoretically expected. But it wasn’t enough to create excess heat.

The allure of cold fusion hasn’t entirely gone away. For example, there are two companies in Japan, Technova Inc. and Clean Planet Inc, which claim to have produced excess heat. Clean Planet Inc has a very impressive roadmap on their website, according to which they’ll complete a model reactor for commercial application next year. There’s just one problem: No one has seen the world-saving machine, and no one has reproduced the results.

The people who still work on cold fusion have renamed it to “Low Energy Nuclear Reactions”, LENR for short. Part of the reason is that “cold” isn’t particularly descriptive. I mean, these devices may be cold compared to the interior of the sun, but they can heat up to some hundred degrees Celsius, and maybe that’s not everybody’s idea of cold. But no doubt the major reason for the rebranding is to get out of the reputation trap. So make no mistake, LENR is cold fusion reborn.

I admit that this doesn’t sound particularly convincing. But I think it’s worth looking a little closer at the details. First of all, there are two separate measurements that cold fusion folks usually look at. That’s the production of decay products from the nuclear fusion, and the production of excess heat.

An experiment that tried to shed light on what might be going on comes from a 2010 paper by a group in the United States. They used a setup very similar to that from Fleischmann and Pons but in addition they directed a pulsed laser at the palladium with specific frequencies. They claimed to see excess power generation for specific pulse frequencies, which suggests that phonon excitations have something to do with it. There’s just one problem: a follow-up experiment failed to replicate the result.

Edmund Storms who has been working on this for decades published a paper in 2016 claiming to have measured excess heat in a device that’s very similar to the original Ponds and Fleischman setup. In this figure you see how the deuterium builds up in the palladium, that’s the red dots, and the amount of power that Storms says he measured.

He claims that the reason that these experiments are difficult to reproduce is that the nuclear reactions happen in appreciable rates only in some regions of the palladium which have specific defects that he calls nano-cracks. These could be caused by the treatment of the metal, so some samples have them and others not, and this is why the experiments sometimes seem to work and sometimes not. At least according to Storms. There’s just one problem: No one’s been able to replicate his findings.

There is also a 2020 paper from the Japanese company, Clean Planet Inc which I already mentioned. They use a somewhat different setup with nanoparticles of certain metals that are surrounded by a gas that contains deuterium. The whole thing is put under pressure and heated. They claim that the resulting temperature increase is higher than you’d expect and that their device generates net power. In this figure you see the measured temperature increase in their experiment with Helium gas and with a gas that contains deuterium. The Helium gas serves as a control. As you see there’s more heating with the deuterium. There’s just one problem: No one’s been able to replicate this finding.

The issue with these heat measurements is that they’re incredibly difficult to verify. For this reason it’s much better to look at the decay products. Those are in and by themselves mysterious. In a typical nuclear fusion reaction, there is a very specific amount of energy that’s released, and so the energy distribution of the decay products is very sharply peaked. In deuterium fusion, the neutrons in particular should have an energy of 2.45 MeV. In those cold fusion reactions, however, they see a fairly broad distribution of neutron energies and at higher energies than expected.

Here is an example. The red bars show the number of deuterium ions as a function of energy, the black ones are the background. As you can see the spectrum looks nowhere like the expected peak at about 2.5 MeV. Something is going on and we don’t know what. Forget saving the world for a moment, it’s much simpler, there’s an observation that we don’t understand.

In a recent paper, a group from MIT has put forward two different hypotheses that could explain why nuclear fusion happens more readily in certain metals than you’d naively assume. One is that there are some unknown nuclear resonances which can become excited and make fusion easier. The other one is that the lattice of the metal facilitates an energy transfer from the deuterium to some of the palladium nuclei. So then you have excited Palladium nuclei and those decay. Since the Palladium nuclei have more decay channels than are typical for fusion outputs, this can explain why the energy distribution looks so weird. There’s just one problem: We don’t know that that’s actually correct.

What are we to make of this? The major reason cold fusion has been discarded as pseudoscience is that most physicist think it can’t possibly be that chemical processes cause nuclear reactions. But I think they overestimate how much we know both about nuclear physics and chemistry.

Nuclear physics is dominated by the strong nuclear force which holds quarks and gluons together so that they form neutrons and protons. The strong nuclear force has the peculiar property that it becomes weaker at high energies. This is called asymptotic freedom. Arvin Ash recently did a great video about the strong nuclear force, so check this out for more details.

The Large Hadron Collider pumps a lot of energy into proton collisions. This is why understanding the strong nuclear force in LHC collisions is quite simple, by which I mean a PhD in particle physics will do. The difficult part comes after the collisions, when the quarks and gluons recombine to protons, neutrons, and other bound states such as pions and rhos and so on. It’s called hadronization, and physicists don’t know how to calculate this. They just extract the properties of these processes from data and parameterize it.

I am telling you this to illustrate that just because we understand the properties of the constituents of atomic nuclei doesn’t mean we understand atoms. We can’t even calculate how quarks and gluons hold together.

Another big gap in our understanding are material properties because we often can’t calculate electron bands. That’s especially true for materials with irregularities that, according to Storms, are relevant for cold fusion. Indeed, if you remember, calculating material properties is one of those questions that physicists want to put on a quantum computer exactly because we can’t currently do the calculation. So, is it possible that there is something going on with the nuclei or electron bands in those metals that we haven’t yet figured out? I think that’s totally possible.

But, let me honest, I find it somewhat suspicious that the power production in cold fusion experiments always just so happens to be very close to the power that goes in. I mean, there isn’t a priori any reason why this should be the case. If there is nuclear fusion going on efficiently, why doesn’t it just blow up the lab and settle the case once and for all?

So, well, I am extremely skeptical that we’ll see a working cold fusion device in the next couple of years. But it seems to me there’s quite convincing evidence that something odd is going on in these devices that deserves further study.

I’m not the only one who thinks so. In the past couple of years, research into cold fusion has received a big funding boost, and that’s already showing results. For example, in 1991, a small group of researchers proposed a method to produce palladium samples that generate excess heat more reliably. And, I hope you’re sitting, research groups at NASA and at the US Navy have recently been able to reproduce those results.

A project at the University of Michigan is trying to reproduce the findings by the Japanese companies. The Department of Energy in the United States just put out a call for research projects on low energy nuclear reactions, and also the European research council has been caught in the act of supporting some cold fusion projects.

I think this is a good development. Cold fusion experiments are small and relatively inexpensive and given the enormous potential, it’s worth the investment. It’s a topic that we’ll certainly talk about again, so if you want to stay up to date, don’t forget to subscribe. Many thanks to Florian Metzler for helping with this video.