One of the biggest challenges in making fusion energy practical is achieving "ignition." This is a state in which the heat produced by fusion reactions keeps the plasma hot enough to sustain itself, without relying on external heating. Reaching it requires maintaining temperatures far higher than those at the center of the Sun, together with sufficient fuel density, for a sustained period. The approach studied most often has been to raise the plasma density first and then apply powerful external heating.
A research team at the Princeton Plasma Physics Laboratory (PPPL), part of the U.S. Department of Energy (DOE), has now published a theoretical model that reverses this order. According to the model, heating a low-density plasma first and then adding fuel to raise the density could reach ignition conditions with less external heating energy than the conventional method. The study was carried out by Luis F. Delgado-Aparicio, Masayuki Ono, and Jonathan E. Menard, and was published in Physical Review Letters on September 10, 2026.
The result is not, however, an experimental one obtained by operating a fusion device. It is a study centered on a zero-dimensional (0D) theoretical model that treats the plasma's overall heat balance, and it also examines the implications for magnetohydrodynamic (MHD) stability in realistic three-dimensional reactor geometries. PPPL explains that current experimental facilities cannot reach the temperature range the model assumes. This is not an announcement that ignition has been achieved. It is a theoretical study showing that rethinking the order of heating and fuel supply on the way to ignition could reduce the energy required.
A "Shortcut" to Fusion: A New Path Revealed by a Theoretical Model
In PPPL's official announcement of the study on October 5, 2026, Delgado-Aparicio uses a mountain-climbing analogy to explain how to reach ignition. In his telling, many companies are trying to reach ignition head-on by investing large amounts of energy, like climbing straight to the summit. But if you go around rather than over the peak and take a lower pass, you can reach the same destination with less energy. The study shows that a similar route exists for heating and fueling a fusion plasma.
The pass in this analogy corresponds to a region known in plasma physics as the "Cordey saddle." It is the point on the energy barrier between a state that requires external heating and a state in which fusion self-heating can sustain burning that can be crossed with relatively low input. The team compared heating and fueling paths through this saddle point using a computational model that accounts for several loss mechanisms. The results suggest that heating a low-density plasma first, and raising the density only after the temperature is high enough, may bring the plasma closer to ignition conditions more efficiently.
It is important to note that this is not a report of extracting net power from an actual fusion reactor. The National Ignition Facility (NIF) at Lawrence Livermore National Laboratory (LLNL) in the United States achieved ignition on December 5, 2022, using laser-driven inertial confinement. That result meant that the energy released by fusion reactions exceeded the laser energy delivered to the fuel target. PPPL's study, by contrast, concerns magnetic confinement, in which strong magnetic fields confine hot plasma in devices such as tokamaks and stellarators. The two approaches differ in both confinement mechanism and operating conditions, and NIF's result does not validate this theoretical model.
According to the paper's record, the study was submitted on April 28, 2026, accepted on August 17, and published on September 10. Although it was peer-reviewed by the American Physical Society, the proposed operating path has not been reproduced in a real device. At this stage, the achievement is a theoretical framework for evaluating in what order fuel should be supplied and at what point heating should be applied in future fusion devices.
What the Lawson Criterion and Q Value Cannot Tell Us: The Path to Ignition
The foundation of fusion research is the "Lawson criterion," derived in the 1950s by British physicist John D. Lawson. To sustain a fusion reaction, the fuel's temperature, density, and energy confinement time must satisfy certain conditions. A representative measure combining them is the triple product , the product of plasma density $n$, temperature $T$, and energy confinement time . In fusion reactor design, whether this triple product meets the ignition condition has long been a key criterion.
Another basic measure is the "Q value," the fusion energy gain factor. It expresses how large the fusion power produced by fusion reactions is relative to the external heating power delivered to the plasma.
$Q=1$ means fusion output equals external heating input, and $Q=5$ means fusion output is five times the external heating input. However, this is a ratio of heating input to fusion output for the plasma, and does not indicate a power plant's overall power consumption or generating efficiency. Ignition, in which the self-heating from alpha particles produced by fusion reactions alone maintains the plasma temperature, requires no external heating and therefore corresponds in theory to the limit .
The Lawson criterion and the Q value are essential for judging what state a fusion reactor should aim for. But these numbers alone do not reveal the most efficient procedure for reaching that state. Raising density first and raising temperature first may involve different energy losses along the way. Even if the final temperature and density are the same, the external heating energy needed to get there is not necessarily the same.
Assuming an ideal fuel mixture of deuterium (D) and tritium (T), the Cordey saddle is said to appear in the region of . In an actual reactor, however, the accumulation of helium produced by burning, impurities entering from the vessel wall, and energy escaping as electromagnetic radiation all alter the plasma's state. With such losses added, the ignition conditions derived for an idealized model cannot be applied as they are. This study seeks to evaluate both the ignition conditions themselves and the path to reach them, taking these real-world losses into account.
A Zero-Dimensional Model Incorporating Four Loss Mechanisms
The study's distinguishing feature is that it integrates, within a single theoretical framework, several physical phenomena that were often examined separately. The paper combines Lawson's power balance, Cordey's accessibility, the marginal ignition ridge, the ignition entry point, and Mills' analysis of thermal runaway to treat fusion plasma ignition conditions systematically. The basis is a zero-dimensional (0D) model that averages out the plasma's internal spatial distribution and calculates the overall energy balance. The team incorporated into this model four loss mechanisms that cannot be ignored in real fusion reactors.
The first is the accumulation of helium ash produced by fusion reactions. When deuterium and tritium fuse, they produce high-energy alpha particles (helium nuclei) and neutrons. Alpha particles play an important role in heating the plasma, but once the helium ions have lost their energy and accumulate in the reactor, they dilute the deuterium and tritium fuel. With the fuel diluted, fewer particles take part in fusion even at the same plasma density, making ignition conditions harder to meet.
The second is light and heavy element impurities entering from the vessel wall. Wall materials exposed to hot plasma can be partly eroded by particle impacts such as sputtering, and their atoms can enter the plasma. The impurities that enter are ionized in the hot environment and cause energy loss through radiation. Heavy elements with high atomic numbers in particular can have a large effect on radiative losses even when only small amounts get in.
The third is energy loss through synchrotron radiation. In magnetic confinement, strong magnetic fields are used to control the motion of charged particles, but electrons emit electromagnetic waves as they spiral around magnetic field lines. This radiation lets energy escape from the plasma, which also affects the heating conditions needed for ignition. In reactors that use strong magnetic fields in particular, the radiative losses that accompany them must be considered, not just the improved confinement performance.
The fourth is growth of heat transport losses as temperature rises. Inside the plasma, heat is carried from the core outward by particle collisions, turbulence, and other processes, and the confined energy is lost. Even if temperature and heating power increase, energy confinement does not necessarily improve in step. The model also takes into account empirical scaling in which energy confinement becomes less favorable as heating power increases.
In PPPL's announcement, co-author Ono explains that ignoring these effects can make a design appear to work without problems, but that including all of them changes the results significantly. In costly fusion experiments, it is important to find design problems before building a device. The aim of this study is likewise to establish a common framework that can evaluate the feasibility of ignition in light of realistic losses, rather than relying only on idealized conditions.
| Item | Ideal deuterium-tritium plasma | This model with four loss mechanisms | Basis / verification status |
|---|---|---|---|
| Analysis assumptions | Assumes pure fuel with no impurities or helium ash | Accounts for helium ash, wall impurities, synchrotron radiation, and heat transport losses | Theoretical analysis in the Physical Review Letters paper |
| Ignition assessment | Centered on threshold conditions such as the triple product | Evaluates the dynamic path to ignition, accounting for losses | Main feature of this theoretical model |
| Cordey saddle | Located near | Shifts toward higher temperature and density; the Q value needed to reach it rises | PPPL announcement and the paper's analysis |
| Required plasma pressure | Calculated from the ideal heat balance | Can nearly double with a trace of tungsten | Calculation published by PPPL; absolute pressure values not given |
| 3D stability | Relationship between ignition conditions and MHD stability not fully treated | Higher required pressure could push the plasma's value beyond stability limits | At the stage of theoretical examination |
| Experimental verification | Related conditions partly verified in existing fusion devices | Operating path through the proposed high-temperature region is unproven | PPPL says current facilities cannot verify it directly |
A New Path to Ignition: Heat First, Add Fuel Later
The model incorporating the four loss mechanisms shows that ignition is harder to achieve than under conventional idealized conditions. As radiative and heat transport losses increase, the marginal ignition ridge, the Cordey saddle, and the ignition entry point all shift toward higher temperature and density. As a result, the plasma pressure needed to achieve ignition also rises. What the team proposes is to deal with these tougher conditions by changing the order of heating and fuel supply.
In conventional operating scenarios, the approach considered has been to bring the plasma density close to its target value first, and then raise the temperature through methods such as neutral beam injection or radio-frequency heating. However, if radiative losses from impurities become large at high density, the heating power needed to raise the plasma temperature also increases. In some cases, losses exceed the energy supplied by heating, making it difficult to reach the target temperature. Because increased radiative cooling can also harm plasma stability, raising density first is not always advantageous.
In the new model, by contrast, a relatively low-density plasma is heated first, and fuel is added once it is hot enough. At low density, some radiative losses associated with particle interactions are easier to suppress, so temperature may be raised efficiently with the same external heating input. Fuel is then supplied, for example by pellet injection, to raise the density, bringing the plasma closer to ignition conditions while increasing fusion self-heating. However, the full operating sequence using pellet injection has not been verified in a real device, and the specific fueling methods and control technologies remain subjects for future study.
In the paper, the team argues that ignition should be reconsidered not as a mere "static threshold condition" but as a "constrained dynamical pathway." The idea is that not only the final temperature and density reached, but also the order in which they are changed, affects the energy required and whether ignition can be reached. In fusion reactor design, in addition to whether target performance can be achieved, how to control the process from the start of operation to ignition is important.
Furthermore, this path is not determined by the plasma's heat balance alone. As the required pressure rises, so does the value, the ratio of plasma thermal pressure to magnetic pressure. If the value is too high, the magnetic field structure confining the plasma risks becoming MHD-unstable. The study thus also shows that the thermodynamic conditions needed for ignition are closely tied to stability limits in realistic three-dimensional reactor geometries.
Trace Tungsten Makes Ignition Conditions Harder
One of the most notable findings of the analysis is the effect of tungsten, a material used for fusion reactor inner walls. In next-generation fusion experimental devices, tungsten is widely adopted as the plasma-facing material because of its high melting point, heat resistance, and resilience against sputtering. But when tungsten, a heavy element with atomic number 74, enters hot plasma, strong radiation causes energy to be lost. In hot environments, some electrons can remain bound to the atoms, and the plasma is cooled through line radiation, bremsstrahlung, and other processes.
According to PPPL, when the fraction of tungsten in the plasma reaches just one part in 10,000, or 0.01%, the plasma pressure needed for ignition nearly doubles. This is a result obtained from the theoretical model, and the absolute pressure values used for comparison are not given in the published materials. PPPL also points out that extending this pressure condition to a three-dimensional reactor geometry, the required pressure could exceed the plasma's stability limit. Even if the temperature and density needed for ignition are secured, a pressure that is too high would make stable magnetic confinement difficult.
This is an important issue for the choice of materials in fusion reactors. Tungsten is excellent as a material that withstands high heat loads, but even a slight intrusion into the plasma could make ignition conditions harder. Raising the heat resistance of the reactor wall is therefore not enough; mechanisms to suppress wall material from entering the plasma, and technologies to control impurities that do enter, are also needed. The analysis highlights the need to design the whole reactor with both materials engineering and plasma physics effects in mind, rather than considering them separately.
On the other hand, the team found that the loss mechanisms that make ignition difficult may also help stabilize the plasma. In fusion plasmas, "thermal runaway," in which self-heating from reactions raises the temperature and thereby makes fusion reactions even more active, can sometimes be a problem. According to this model, however, heat transport, impurity radiation, synchrotron radiation, and the degradation of confinement performance as heating power increases also work to suppress such temperature rises. The paper indicates that negative feedback from multiple loss mechanisms may suppress the growth of thermal runaway and contribute to maintaining a steady burning state.
This does not mean, however, that greater losses make a fusion reactor more stable. If losses are too large, meeting ignition conditions becomes difficult in the first place. What matters is the possibility that a stable balance can be established between the temperature rise from self-heating and energy loss. Because a fusion reactor holds only limited fuel in its core, and fusion reactions stop if the conditions needed to sustain them break down, it differs in nature from the runaway chain reaction possible in a fission reactor. Still, the thermal instability of plasma examined here must be distinguished from the safety and operational risks of the device as a whole.
Technologies to Improve Ignition Conditions, and the Barrier of Experimental Verification
The team also mentions ways to ease ignition conditions made harsher by real-world losses. One is to coat the inner wall of the vacuum vessel with a thin film of liquid lithium. This technology, which PPPL has long studied, is expected to suppress tungsten from entering the plasma and improve heat confinement. However, this study does not newly demonstrate the ignition performance of reactors using liquid lithium; it is positioned as one of the improvement measures examined in existing research.
Another is "spin-polarized fuel." This technology controls the orientation of the spin, a quantum-mechanical angular momentum, of deuterium and tritium nuclei to make fusion reactions more likely. Properly aligning the spin states is thought to increase the fusion reaction cross-section at the same temperature and density. If the reaction rate can be raised, the conditions needed for ignition may be eased.
However, what the paper shows is that improving energy confinement performance and introducing spin-polarized fuel can partially bring degraded ignition conditions back toward the level of ideal deuterium-tritium plasma. The authors themselves carefully say that these improvements "partially reopen" the range of conditions in which ignition is possible. Losses from impurities and radiation cannot be eliminated entirely, so these should be regarded as means of improving realistic operating conditions.
Significant limitations remain in this study. According to PPPL, the high-temperature region near the Cordey saddle, through which the proposed heat-first path passes, cannot be reached by current experimental devices. It therefore remains unconfirmed whether external heating energy can be reduced as the model predicts, whether fueling after heating can be controlled stably, and whether the effects of real impurities and heat transport will match the calculations. In particular, whether plasma stability can be maintained while temperature and density are being changed is an important verification issue for practical use.
In PPPL's announcement, co-author Menard acknowledges that further research is needed while saying the results give a clear direction for future work. The team plans to verify the heat-first path through simulations that reproduce plasma behavior in greater detail on computers. In the paper, the integrated model built here is positioned as a foundation for evaluating the accessibility, stability, and performance of ignition in next-generation fusion devices.
This study does not directly guarantee an earlier arrival of fusion power generation or lower power generation costs. Even so, the perspective of optimizing not only the temperature and density needed to meet ignition conditions but also the operating path that leads there is significant. Going forward, more detailed simulations and experiments will be needed to confirm whether this theoretical path is also effective in actual fusion devices. For the practical use of fusion reactors, research on operational control, namely how to reach ignition conditions using limited heating energy, will become important alongside efforts to improve device performance.