Conversation with Anders Sandberg on antimatter weapons

News that the US Airforce is pursuing an antimatter weapons program hit the media last week. Concerned that this theoretical technology could represent a new existential risk to our species, I opened a dialogue with Anders Sandberg to get his opinions on the subject. Here's our conversation:

Anders:

I wonder how serious this risk is. I have played around with antimatter weapon scenarios a bit, and they do not seem to be much worse than nuclear weapons. They still ought to be beside them on the list, but they do not seem to do anything truly new.

In the near future they are utterly too expensive to be useful, and they require major installations to build. This places them in the same fairly easily controlled category as nuclear weapons. Since containment is also active, proliferation is unlikely - simply too hard to keep stable.

Things change a bit if we move to space. There one can build "antimatter distilleries" orbiting the sun, accumulating antimatter or producing it through solar-powered accelerators. Everything scales up, and now one can make sizeable amounts. Production is still very visible, and fairly easily disrupted.

Antimatter becomes a serious problem if a stable (especially a passive) containment is constructed (or it can be produced in such a form). Penning traps are too small to be a threat, but once you have stably magnetically levitated anti-hydrogen or heavier, you have a potential problem with proliferation. Amat bombs could (if the containment is small) be sneaked around relatively easily and be hard to keep track of. Again, I don't think this is worse than suitcase nuclear proliferation. You get a very unstable and risky situation, but not the end of the world (Charles Stross description of the situation on the post-singularity Earth in his latest novels comes to mind).

Big antimatter detonations could be used to sterlize planetary surfaces if spread across the atmosphere. The ambiplasma fireball is going to radiate a very nasty blackbody spectrum over a sustained amount of time (minutes?) and likely produce lots of short-lived biologically active isotopes. But this kind of sterilization seems to require plenty of preparation (making the antimatter in significant amounts) and deliberate use, not just accidents. Most likely a Shriek rather than an accidental Bang.

Just some thoughts.

Me:

Thanks for the clarification and insight, Anders. It certainly doesn't seem like the kind of weapon that would be utilized by nihilist groups/individuals or even aggressive state actors.

Anders:

Maybe aggressive states would find uses for it, but beyond a certain size bigger explosions are not more destructive (most of the explosive power ends up in space). Maybe antimatter would be good for secretive space warfare or redirecting asteroids at enemies, but that seems a relatively low-level nastiness.

Me:

That being said, could a catastrophic accident arise from the development of such a weapon or fuel source (e.g. antimatter catalyzed nuclear pulse propulsion)?

Anders:

I'm thinking possibly, but only in the form a localized disaster. A Shriek at Worst.

Hmm, to accelerate a starship to 0.3 c you need about 2.5% antimatter if you have perfect conversion of energy into kinetic energy. A million ton ship would need 25,000 ton antimatter, which would indeed be a bad day if it all blew up. About 976,000 gigatons bad. (assuming I calculated right). This kind of space opera starship would indeed be a dinosaur killer.

Going down to a mere 0.1 c gives a requirement of just 0.25% antimatter, and 0.01 c 2.5e-3% - much safer.

Me:

Although, how do you know how long the ambiplasma burst will linger? It's my understanding that they can be relatively long-lived, as the component particles and antiparticles are too hot and too low-density to annihilate with each other rapidly. If the ambiplasma fireball lingered for a protracted period, that could be bad.

Anders:

I tried to calculate this during my recent trip to Brussels, as I thought my earlier answer was a bit too handwavy, but the exact answer seems hard. The mean free path depends on the density, but the density will depend on how the initial fireball expands. This will in turn depend on how opaque to gamma rays the air is and how it is heated; my general impression is that there will be a high density shockwave accelerating outwards with much antimatter reactions (as fast moving antimatter reaches it from inside) and driven by internal gamma-rays and a low density ambiplasma with more antimatter inside. Given that average particle speeds will be hundreds of km/s scattering will be very intense unless the gas is very dilute. See this article.

But to do this correctly I would likely need a nuclear fireball simulator. Maybe we should ask Sandia labs :-)

Given these ponderings, I'm revising my estimates of ambiplasma longevity in an atmosphere downwards. You get a nasty fireball, but it won't last *much* longer than a nuclear one.

Me:

Don't antimatter explosions produce copious amounts of gamma rays?

Anders:

Yup. That is the primary effect of quark-antiquark and lepton-antilepton interactions. But don't forget that the reactions also cause cascades of very hot nuclei, protons, neutrons and leptons - you get all kinds of rays.

Me:

Gamma rays are very dangerous and are the primary source of the radioactive fallout of nuclear weapons. I recognize that the effects would be immediate (unlike radioactive isotopes which persist), but I wonder what the effect of that would be on biological organisms.

Anders:

Gamma rays in themselves do not cause fallout. They just hurt organic matter, but have to be very strong to disrupt atomic nuclei. In this case they will mainly heat up the fireball, and get scattered into (still dangerous) x-rays.

Actually, an antimatter bomb would be dirty since any atom getting hit by the antimatter would be transmuted. In air that would likely mean that you got lots of light nuclei. This isn't that bad, because most of them decay fairly quickly before they can get integrated into an organism - but *lots* of C14 and tritium could be troublesome. Unfortunately, a lot of highly accelerated alpha particles and others will slam into heavier nuclei and leave behind nastier isotopes, especially if it is a ground-burst.

Me:

Oh, one final thought, is there any connection between gamma ray bursters and antimatter explosions? Do antimatter explosions happen naturally in the Universe?

Anders:

Not to my knowledge. Antimatter detonations would have a characteristic spectrum (like at 0.5 MeV for electron-positron annihilation), and it is not seen in GRBs or elsewhere. No risk for antimatter rocks hitting us.