TAE Technologies aims to demonstrate the value of proton-boron fusion fuel. Image: TAE Technologies

The California-based company TAE Technologies recently announced the results of a series of fusion experiments conducted in collaboration with Japan’s National Institute for Fusion Science (NIFS). The findings have been published in the journal Nature Communications.

In these experiments, alpha particles have been observed that are attributed to proton-boron fusion reactions. The outcome has been characterized as the first observation of proton-boron fusion in a magnetic confinement configuration – and as such has been hailed by some press outlets as another major fusion breakthrough.

But how significant are these results really, and how to situate them in the current state of fusion research overall?

Announcements of breakthroughs in fusion research appear on an almost monthly basis now. More than 50 fusion companies exist today – and almost as many different approaches toward commercial fusion are pursued by these firms. This article seeks to help interested readers to contextualize such announcements and to gauge their meaning and implications.

In doing so, it will discuss some of the critical tradeoffs and uncertainties that prevail in the emerging fusion industry. This specifically pertains to the choice of fuel (proton-boron versus deuterium-tritium) and the way energy is delivered to the fuel (via lasers versus particle beams combined with electromagnetic heating).

The TAE experiments

In the TAE experiments, a hydrogen plasma is heated and suspended in a vacuum chamber and then hit with beams of boron nuclei. The suspension of the plasma is conducted through a complex magnetic field generated by a set of coils that surround the chamber — making this a magnetic confinement approach to fusion. Researchers then observed alpha particles from particle detectors in the chamber.

The chamber used in these experiments is known as the Large Helical Device (LHD), a roughly four-meter-wide donut-shaped vessel of the stellarator type that was built in Japan’s Gifu Prefecture during the 1990s. The LHD is a fusion research device designed to explore the basic physics of magnetic confinement fusion.

Researchers say trillions of proton-boron fusion reactions occurred every second within the twisted intestines of Japan’s Large Helical Device. Image: National Institute for Fusion Science, Japan

Even though TAE’s fusion approach is ultimately based on a different cylindrical vessel, the local conditions where the hydrogen plasma and the boron beams meet exhibit many similarities, making this a suitable research configuration for TAE.

The reported alpha particles are energetic helium nuclei consisting of two protons and two neutrons. The nuclei in the proton-boron fuel are simply hydrogen nuclei (i.e. protons), as found in water, boron nuclei and many household cleaning supplies.

When a proton and boron nucleus collide at high enough energy, fusion can take place resulting in a single nucleus, which soon after disintegrates into three alpha particles (in the large majority of cases, there are some exceptions). The nuclear energy that gets released in the process gets carried away by these alpha particles and can be subsequently harnessed.

The research team behind the TAE experiments estimates that per second on the order of 1014 such fusion reactions have been induced in their experimental campaign. This corresponds to released energy on the order of 100 joules, or 0.00003 kWh. From a research perspective, this is a significant number but it is still many orders of magnitude too low to make up for the energetic cost of creating the required fusion conditions.

With eyes on commercial fusion, key questions are: How can energy yields be increased beyond these results? And: How do competing approaches stack up in comparison?

Fuel of choice?

A major point of ongoing discussion in the fusion community is the optimal choice of fuel. And, in particular, the question of whether proton-boron is indeed a feasible fuel choice or whether the traditionally more widely pursued deuterium-tritium combination is more likely to become prevalent for fusion energy at scale.

The intricacies of nuclear physics make it such that energy release results from the fusion of light nuclei (and from the fission of heavy nuclei, as exploited in fission power plants). As a rule of thumb, the higher up in the periodic table certain elements are, the harder fusion becomes. This is why hydrogen as the lightest of all elements, and its isotopes deuterium and tritium, are common candidates for fusion fuel – and boron, with its five protons, much less so.

This is reflected in responses to the 2022 Fusion Industry Association survey: among 33 companies that participated in the survey, 22 pursue deuterium-tritium fusion. However, there are tradeoffs to be considered here. While proton-boron fusion is harder to be induced, it also boasts certain advantages over deuterium-tritium fusion, and thus six of the 33 companies pursue it.

The downsides of the deuterium-tritium fuel combination are as follows: (1) tritium is radioactive, which means that it needs to be actively managed, and it needs to be produced on a regular basis (as it decays with a half-life of about 12 years); (2) The deuterium-tritium reaction yields a large number of high-energy neutrons which can damage reactor interiors and surrounding equipment, and make some exposed materials radioactive.

Effectively this means higher costs in handling fuel and greater regulatory demands, the need to produce tritium on an ongoing basis and regular replacement of certain reactor components.

Presently, most tritium in circulation comes from Canadian fission reactors, where it is generated as a byproduct and extracted. However, for fusion at scale, much larger quantities are needed.

This is why most tritium-based fusion companies pursue schemes to generate tritium as a byproduct of their fusion reactions. However, much uncertainty prevails as to what quantities of tritium can actually be produced this way.

Given these drawbacks of deuterium-tritium fuel, the question is how the challenges of proton-boron stack up in comparison.

A common concern about proton-boron fusion is that temperatures in excess of 1 billion degrees Celsius are needed. Such numbers can be misleading, however.

(Left) 3D CAD model showing the LHD vacuum vessel with cut-away view of heliotron plasma. (Right) CAD image showing calculated alpha particle trajectories (green curves) reaching the PIPS detector near the LHD separatrix, a portion of the last closed flux surface (tan), and the PIPS detector, located below the plasma. Images: Nature

In a plasma in thermal equilibrium, a given temperature number corresponds to a broad distribution of particle energies, whereby much of the imparted heating energy ends up in an ineffectively low part of the energy spectrum.

In a beam, by contrast, energy can be imparted onto particles in a more controlled manner, allowing researchers to target regions of the energy spectrum that are more desirable for inducing fusion reactions.

Moreover, at higher particle energies – above about 400 keV – the fusion probability of proton-boron fuel exceeds that of deuterium-tritium fuel. What is helpful then is for energy to the fuel to be deployed in such a way that much of it triggers actual fusion reactions rather than getting lost, e.g. as heat escaping into the environment. Different concepts for doing so have been proposed in the research communities but even experts tend to disagree on how feasible they are.

From a big-picture perspective, there are those in the fusion community that consider the challenges of deuterium-tritium fusion to be more manageable than the challenges of proton-boron fusion; and those who arrive at the opposite conclusion. TAE falls into the camp of proton-boron proponents, even though the company has also worked with deuterium-tritium fuels as a stepping stone.

Proton-boron preference

Even among proton-boron proponents themselves, though, there are differences in the assessment of alternative approaches.

TAE argues that a magnetic confinement approach – along the lines of the described experiments – is most likely to result in fusion at scale. But how to close the gap between the roughly 1014 fusion reactions per second that have now been reported and the orders of magnitude higher number needed for energy production?

The TAE authors argue that a combination of known mechanisms that involve stirring and concentrating particles in a plasma could be used to greatly increase the efficiency and yields of their system. However, before such sought plasma effects are experimentally demonstrated in real-world scenarios, much uncertainty remains as to their feasibility.

Other researchers are skeptical whether the above-mentioned mechanisms will be sufficient to cross the bar for energy production. They instead focus on collections of effects that can occur in systems where lasers are used to accelerate fusion reactants. In fact, much of the experimental proton-boron fusion research to date has employed laser-driven approaches.

Over the years, alpha particle yields on the order of 1010 particles per laser shot have been achieved this way. Since the laser shots used in these experiments last only a fraction of a second, larger absolute values of generated particles could be achieved by increasing the laser repetition rates.

However, this doesn’t change the fact that, until now, each laser shot contains more than 10,000 times the energy released in the induced fusion reactions.

A laser fusion microexplosion. Photo: US Department of Energy

Here, too, orders of magnitude of yield increases are needed to expand on such experimental results toward commercial fusion energy. Different researchers (e.g. Belloni, Ruhl) have pointed at energy transfer mechanisms between energetic nuclei in a plasma as helpful to that end.

However, as in the case above, some uncertainty remains until the magnitude of such purported effects is conclusively demonstrated in dedicated experiments.


While substantial progress has been and is being made in many areas of fusion science, much uncertainty prevails with respect to alternative design choices. This includes the choice of fuel as well as different energy delivery and confinement mechanisms.

The uncertainty here is of such a nature that even experts can disagree fundamentally in the assessment of risks and opportunities of competing alternatives. Moreover, few scientists – even highly decorated ones – possess uniform depth of knowledge and experience across all of the many relevant scientific subfields. This further adds to the inevitable subjectivity of assessments that inform design and investment decisions.

The TAE results represent rigorously conducted and carefully documented incremental progress toward commercial fusion energy along one of the various trajectories pursued by fusion researchers.

They are exciting news for those who are convinced that the opportunities of proton-boron fusion outweigh the challenges of deuterium-tritium fusion, and that beam injection into a plasma may be the most likely-to-succeed approach to commercial fusion.

However, it is too early to predict which fusion approaches will turn out to be most suitable for large-scale dissemination of fusion energy production. Many of the dozens of pursued trajectories will not make it beyond their prototype phases.

On the other hand, there will likely not be just one silver bullet – but rather a set of different reactor types and fuels, each exhibiting their own pros and cons that match different use cases and regulatory regimes.