Aneutronic fusion is any form of fusion power in which neutrons carry no more than 1% of the total released energy. The most-studied fusion reactions release up to 80% of their energy in neutrons. Successful aneutronic fusion would greatly reduce problems associated with neutron radiation such as ionizing damage, neutron activation and requirements for biological shielding, remote handling and safety.
Some proponents see a potential for dramatic cost reductions by converting energy directly to electricity. However, the conditions required to harness aneutronic fusion are much more extreme than those required for the conventional deuterium–tritium (D-T) nuclear fuel cycle.
Multiple fusion reactions have no neutrons as products on any of their branches. Those with the largest cross sections are these:
|Deuterium–helium-3||2D||+||3He||→||4He||+||1p||+ 18.3 MeV|
|Deuterium–lithium-6||2D||+||6Li||→||2||4He||+ 22.4 MeV|
|Proton–lithium-6||1p||+||6Li||→||4He||+||3He||+ 4.0 MeV|
|Helium-3–lithium||3He||+||6Li||→||2||4He||+||1p||+ 16.9 MeV|
|Helium-3-helium-3||3He||+||3He||→||4He||+||2 1p||+ 12.86 MeV|
|Proton–lithium-7||1p||+||7Li||→||2||4He||+ 17.2 MeV|
|Proton–boron||1p||+||11B||→||3||4He||+ 8.7 MeV|
|Proton–nitrogen||1p||+||15N||→||12C||+||4He||+ 5.0 MeV|
The deuterium reactions produce some neutrons with D-D side reactions. Although these can be minimized by running hot and deuterium-lean, the fraction of energy released as neutrons is probably several percent, so that these fuel cycles, although neutron-poor, do not meet the 1% threshold. See Helium-3. These reactions also suffer from the 3He fuel availability problem, as discussed below.
The proton-lithium-6, helium-3-lithium, and helium-3-helium-3 reaction rates are not particularly high in a thermal plasma. When treated as a chain, however, they offer the possibility of enhanced reactivity due to a non-thermal distribution. The product 3He from the Proton-lithium-6 reaction could participate in the second reaction before thermalizing, and the product p from helium-3-lithium could participate in the former before thermalizing. Unfortunately, detailed analyses do not show sufficient reactivity enhancement to overcome the inherently low cross section.
The pure 3He reaction suffers from a fuel-availability problem. 3He occurs in only minuscule amounts naturally on Earth, so it would either have to be bred from neutron reactions (counteracting the potential advantage of aneutronic fusion), or mined from extraterrestrial sources. The top several meters of the surface of the Moon is relatively rich in 3He on the order of 0.01 parts per million by weight, but mining this resource and returning it to Earth would be relatively difficult and expensive. 3He could in principle be recovered from the atmospheres of the gas giant planets, Jupiter, Saturn, Neptune and Uranus, but this would be even more challenging. The amount of fuel needed for large-scale applications can also be put in terms of total consumption: according to the US Energy Information Administration, "Electricity consumption by 107 million U.S. households in 2001 totaled 1,140 billion kW·h" (1.14×1015 W·h). Again assuming 100% conversion efficiency, 6.7 tonnes per year of helium-3 would be required for that segment of the energy demand of the United States, 15 to 20 tonnes per year given a more realistic end-to-end conversion efficiency. Extracting that amount of pure helium-3 would entail processing 2 billion tonnes of lunar material per year, even assuming a recovery rate of 100%.
The fusion of a boron nucleus with a proton uses one laser to create a boron-11 plasma and another to create a stream of protons that smash into the plasma, producing slow-moving helium particles but no neutrons. The laser-generated proton beam produces a tenfold increase of boron fusion because protons and boron nuclei collide directly. Earlier methods used a solid boron target, "protected" by its electrons, which reduced the fusion rate. 
The plasma lasts about one-billionth of a second, requiring the pulse of protons, which lasts one-trillionth of a second, to be precisely synchronized. The proton beam is preceded by a beam of electrons, generated by the same laser, that pushes away electrons in the boron plasma, allowing the protons more of a chance to collide with the boron nuclei and initiate fusion.
Detailed calculations show that at least 0.1% of the reactions in a thermal p–11B plasma would produce neutrons, and the energy of these neutrons would account for less than 0.2% of the total energy released.
These neutrons come primarily from the reaction
The reaction itself produces only 157 keV, but the neutron will carry a large fraction of the alpha energy, which will be close to Efusion/3 = 2.9 MeV. Another significant source of neutrons is the reaction
These neutrons are less energetic, with an energy comparable to the fuel temperature. In addition, 11C itself is radioactive, but will decay to negligible levels within hours as its half life is only 20 minutes.
Since these reactions involve the reactants and products of the primary fusion reaction, it would be difficult to further lower the neutron production by a significant fraction. A clever magnetic confinement scheme could in principle suppress the first reaction by extracting the alphas as soon as they are created, but then their energy would not be available to keep the plasma hot. The second reaction could in principle be suppressed relative to the desired fusion by removing the high energy tail of the ion distribution, but this would probably be prohibited by the power required to prevent the distribution from thermalizing.
with a branching probability relative to the primary fusion reaction of about 10−4.
With careful design, it should be possible to reduce the occupational dose of both neutron and gamma radiation to operators to a negligible level. The primary components of the shielding would be water to moderate the fast neutrons, boron to absorb the moderated neutrons and metal to absorb X-rays. The total thickness needed should be about a meter, mostly water.
Many challenges remain prior to the commercialization of aneutronic processes.
The large majority of fusion research has gone toward D-T fusion, because the technical challenges of proton–boron fusion are so formidable. Proton–boron fusion requires ion energies or temperatures almost ten times higher than those for D-T fusion. For any given density of the reacting nuclei, the reaction rate for proton-boron achieves its peak rate at around 600 keV (6.6 billion degrees Celsius or 6.6 gigakelvins)[better source needed] while D-T has a peak at around 66 keV (765 million degrees Celsius). For pressure-limited confinement concepts, optimum operating temperatures are about 5 times lower, but the ratio is still roughly ten-to-one.
The peak reaction rate of p–11B is only one third that for D-T, requiring better plasma confinement. Confinement is usually characterized by the time τ the energy must be retained so that the fusion power released exceeds the power required to heat the plasma. Various requirements can be derived, most commonly the product of the density, nτ, and the product with the pressure nTτ, both of which are called the Lawson criterion. The nτ required for p–11B is 45 times higher than that for D-T. The nTτ required is 500 times higher. (See also neutronicity, confinement requirement, and power density.) Since the confinement properties of conventional fusion approaches, such as the tokamak and laser pellet fusion are marginal, most aneutronic proposals use radically different confinement concepts.
In most fusion plasmas, bremsstrahlung radiation is a major energy loss channel. (See also bremsstrahlung losses in quasineutral, isotropic plasmas.) For the p–11B reaction, some calculations indicate that the bremsstrahlung power will be at least 1.74 times larger than the fusion power. The corresponding ratio for the 3He-3He reaction is only slightly more favorable at 1.39. This is not applicable to non-neutral plasmas, and different in anisotropic plasmas.
In conventional reactor designs, whether based on magnetic or inertial confinement, the bremsstrahlung can easily escape the plasma and is considered a pure energy loss term. The outlook would be more favorable if the plasma could reabsorb the radiation. Absorption occurs primarily via Thomson scattering on the electrons, which has a total cross section of σT = 6.65×10−29 m². In a 50–50 D-T mixture this corresponds to a range of 6.3 g/cm². This is considerably higher than the Lawson criterion of ρR > 1 g/cm², which is already difficult to attain, but might be achievable in inertial confinement systems.
In megatesla magnetic fields a quantum mechanical effect may suppress energy transfer from the ions to the electrons. According to one calculation, bremsstrahlung losses could be reduced to half the fusion power or less. In a strong magnetic field cyclotron radiation is even larger than the bremsstrahlung. In a megatesla field, an electron would lose its energy to cyclotron radiation in a few picoseconds if the radiation could escape. However, in a sufficiently dense plasma (ne > 2.5×1030 m−3, a density greater than that of a solid), the cyclotron frequency is less than twice the plasma frequency. In this well-known case, the cyclotron radiation is trapped inside the plasmoid and cannot escape, except from a very thin surface layer.
While megatesla fields have not yet been achieved, fields of 0.3 megatesla have been produced with high intensity lasers, and fields of 0.02–0.04 megatesla have been observed with the dense plasma focus device.
At much higher densities (ne > 6.7×1034 m−3), the electrons will be Fermi degenerate, which suppresses bremsstrahlung losses, both directly and by reducing energy transfer from the ions to the electrons. If necessary conditions can be attained, net energy production from p–11B or D–3He fuel may be possible. The probability of a feasible reactor based solely on this effect remains low, however, because the gain is predicted to be less than 20, while more than 200 is usually considered to be necessary. (Other effects might improve the gain substantially.)
In every published fusion power plant design, the part of the plant that produces the fusion reactions is much more expensive than the part that converts the nuclear power to electricity. In that case, as indeed in most power systems, power density is an important characteristic. Doubling power density at least halves the cost of electricity. In addition, the confinement time required depends on the power density.
It is, however, not trivial to compare the power density produced by different fusion fuel cycles. The case most favorable to p–11B relative to D-T fuel is a (hypothetical) confinement device that only works well at ion temperatures above about 400 keV, in which the reaction rate parameter <σv> is equal for the two fuels, and that runs with low electron temperature. p–11B does not require as long a confinement time because the energy of its charged products is two and a half times higher than that for D-T. However, relaxing these assumptions, for example by considering hot electrons, by allowing the D-T reaction to run at a lower temperature or by including the energy of the neutrons in the calculation shifts the power density advantage to D-T.
The most common assumption is to compare power densities at the same pressure, choosing the ion temperature for each reaction to maximize power density, and with the electron temperature equal to the ion temperature. Although confinement schemes can be and sometimes are limited by other factors, most well-investigated schemes have some kind of pressure limit. Under these assumptions, the power density for p–11B is about 2,100 times smaller than that for D-T. Using cold electrons lowers the ratio to about 700. These numbers are another indication that aneutronic fusion power is not possible with mainline confinement concepts.
None of these efforts has yet tested its device with hydrogen–boron fuel, so the anticipated performance is based on extrapolating from theory, experimental results with other fuels and from simulations.
Aneutronic fusion produces energy in the form of charged particles instead of neutrons. This means that energy from aneutronic fusion could be captured using direct conversion instead of the steam cycle that is used for neutrons. Direct conversion techniques can either be inductive, based on changes in magnetic fields, electrostatic, based on pitting charged particles against an electric field or, photoelectric, in which light energy is captured. In a pulsed mode, inductive techniques could be used.
Electrostatic direct conversion uses charged particles’ motion to create voltage. This voltage drives electricity in a wire. This becomes electrical power, the reverse of most phenomena that use a voltage to put a particle in motion. Direct energy conversion does the opposite. It uses a particle's motion to produce a voltage. It has been described as a linear accelerator running backwards. An early supporter of this method was Richard F. Post at Lawrence Livermore. He proposed to capture the kinetic energy of charged particles as they were exhausted from a fusion reactor and convert this into voltage to drive current in a wire. Post helped develop the theoretical underpinnings of direct conversion, which was later demonstrated by Barr and Moir. They demonstrated a 48 percent energy capture efficiency on the Tandem Mirror Experiment in 1981.
Aneutronic fusion loses much of its energy as light. This energy results from the acceleration and deceleration of charged particles. These speed changes can be caused by Bremsstrahlung radiation or cyclotron radiation or synchrotron radiation or electric field interactions. The radiation can be estimated using the Larmor formula and comes in the X-ray, IR, UV and visible spectrum. Some of the energy radiated as X-rays may be converted directly to electricity. Because of the photoelectric effect, X-rays passing through an array of conducting foils transfer some of their energy to electrons, which can then be captured electrostatically. Since X-rays can go through far greater material thickness than electrons, many hundreds or thousands of layers are needed to absorb the X-rays.