On March 24, 2004, Eglin Air Force Base Munitions Directorate official Kenneth Edwards spoke at the NASA Institute for Advanced Concepts During the speech, Edwards ostensibly emphasized a potential property of positron weaponry, a type of antimatter weaponry: Unlike thermonuclear weaponry, positron weaponry would leave behind "no nuclear residue", such as the nuclear fallout generated by the nuclear fission reactions which power nuclear weapons. According to an article in San Francisco Chronicle, Edwards has granted funding specifically for positron weapons technology development, focusing research on ways to store positrons for long periods of time, a significant technical and scientific difficulty.
There is considerable skepticism within the physics community about the viability of antimatter weapons. According to an article on the website of the CERN laboratories, which produces antimatter on a regular basis, "There is no possibility to make antimatter bombs for the same reason you cannot use it to store energy: we can't accumulate enough of it at high enough density. (...) If we could assemble all the antimatter we've ever made at CERN and annihilate it with matter, we would have enough energy to light a single electric light bulb for a few minutes."
In 2004, the annual production of antiprotons at the Antiproton Decelerator facility of CERN was several picograms at a cost of $20 million. Thus, at the current level of production, one gram of antimatter would cost $100 quadrillion and would take 100 billion years to produce. While since the first creation of antiprotons, production rates have increased nearly geometrically, it is difficult to say whether antimatter production will ever be rapid or cheap enough to enable military uses. Physical laws such as the small cross-section of antiproton production in high-energy nuclear collisions make it difficult and perhaps impossible to drastically improve the production efficiency of antimatter.
The second problem is the containment of antimatter. Antimatter annihilates with regular matter on contact, so it would be necessary to prevent contact, for example by producing antimatter in the form of solid charged or magnetized particles, and suspending them using electromagnetic fields in near-perfect vacuum. Another, more hypothetical method is the storage of antiprotons inside fullerenes. The negatively charged antiprotons would repel the electron cloud around the sphere of carbon, so they could not get near enough to the normal protons to annihilate with them.
In order to achieve compactness given macroscopic weight, the overall electric charge of the antimatter weapon core would have to be very small compared to the number of particles. For example, it is not feasible to construct a weapon using positrons alone, due to their mutual repulsion. The antimatter weapon core would have to consist primarily of neutral antiparticles. Extremely small amounts of antihydrogen have been produced in laboratories, but containing them (by cooling them to temperatures of several millikelvins and trapping them in a Penning trap) is extremely difficult. And even if these proposed experiments were successful, they would only trap several antihydrogen atoms for research purposes, far too few for weapons or spacecraft propulsion. Heavier antimatter atoms have yet to be produced.
The difficulty of preventing accidental detonation of an antimatter weapon may be contrasted with that of a nuclear weapon. In an antimatter weapon, any failure of containment would immediately result in energy release, which would probably further damage the containment system and lead to the release of all of the antimatter material, causing the weapon to explode at some very substantial fraction of its full yield. By contrast, a modern nuclear weapon will explode with a significant yield if and only if the chemical explosive triggers are fired at precisely the right sequence at the right time, and a neutron source is triggered at exactly the right time. In short, an antimatter weapon would have to be actively kept from exploding; a nuclear weapon will not explode unless active measures are taken to make it do so.
The overall structure of energy output from an antimatter bomb is highly dependent on the amount of regular matter in the area surrounding the bomb. If the bomb is shielded by sufficient amounts of matter, the gamma rays are absorbed and the pions slow down before decaying. Part of the kinetic energy is thus transferred to the surrounding atoms, which heat up.
In any practical form however, the weapon could not simply be a ball of antimatter floating in space. There would have to be a significant amount of supporting hardware surrounding the antimatter. Also, in order to maximize the power of the bomb, it would be designed to mix the antimatter with matter in the least amount of time. The effect of a large antimatter bomb would likely be similar to that of a nuclear explosion of similar size. The reacting antimatter would release about half of its energy in a form immediately available to the environment, superheating the casing and components of the bomb and the surrounding air, and turning it into an ultrahot plasma which then emits blackbody radiation in the full EM spectrum. A quantity as small as a kilogram of antimatter would release 1.8×1017 J (180 petajoules) of energy. Given that roughly half the energy will escape as non interacting neutrinos, that gives 90 petajoules of combined blast and EM radiation, or the rough equivalent of a 20 megaton thermonuclear bomb.
Antimatter catalyzed nuclear pulse propulsion proposes the use of antimatter as a "trigger" to initiate small nuclear explosions; the explosions provide thrust to a spacecraft. The same technology could theoretically be used to make very small and possibly "fission-free" (very low nuclear fallout) weapon (see Pure fusion weapon). Antimatter catalysed weapons could be more discriminate and result in less long-term contamination than conventional nuclear weapons, and their use might therefore be more politically acceptable.
Igniting fusion fuel requires at least a few kilojoules of energy (for laser induced fast ignition of fuel precompressed by a z-pinch), which corresponds to around 10−13 gram of antimatter, or 1011 anti-hydrogen atoms. Fuel compressed by high explosives could be ignited using around 1018 protons to produce a weapon with a one kiloton yield. These quantities are clearly more feasible than those required for "pure" antimatter weapons, but the technical barriers to producing and storing even small amounts of antimatter remain formidable.