# Antimatter Storage

How would it be possible to store a large amount of antimatter (say, 1000000 kg) for a long amount of time (relatively speaking, like around 100 years, or long enough to travel to other systems in interstellar space)? I know that this would be impossible with our current technology but is it be theoretically possible?

• Are you more concerned about storage than production? Have you calculated how much energy you’ll need to produce this amount? – G. Smith Dec 16 '19 at 17:18
• If you make the whole spaceship out of antimatter and crew it with anti-humans, it might be easier... – Jon Custer Dec 16 '19 at 17:27
• Not really, I just wanted to know how you would store the antimatter – user248541 Dec 16 '19 at 17:35
• FWIW, 1000 tonnes of antimatter + 1000 tonnes of matter will take 2.25 tonnes of payload about 27 lightyears, assuming you accelerate at 1g, and you have a perfect engine. See math.ucr.edu/home/baez/physics/Relativity/SR/Rocket/rocket.html – PM 2Ring Dec 16 '19 at 17:45
• As for storage, see en.wikipedia.org/wiki/Penning_trap although that may not be so practical for large quantities of antimatter. ;) – PM 2Ring Dec 16 '19 at 17:49

Theoretically it is entirely possible. Just place a block of anti-tungsten in a perfect vacuum and hold it in place using magnetic fields (it is paramagnetic)

Is this actually doable according to known physics? There are a few problems: (1) evaporation from the antimatter or container walls causing drifting atoms, and (2) making anti-tungsten. I will ignore (2) which is of course a magnificent can of worms and instead deal with (1).

The vapour pressure of tungsten is $$\approx 10^{7.933-45087/ T }$$ atmospheres. Already at 300K this produces a pressure so low that one should not expect any loose atoms, but since we get to work in theory-land we can of course make $$T$$ as low as we want. Same thing for the container walls. We can add turbomolecular pumps to maintain low pressure.

Still, in real 3K cryogenic vacuum systems 6.7 fPa has been achieved, corresponding to 100 particles per cubic centimetre (indeed, just having the system open to interstellar space would reduce this density by a factor of $$\approx 100$$). If we assume hydrogen molecules their RMS speed is $$v_{rms}=\sqrt{\frac{3RT}{m}}=192.6778$$ m/s. So the surface of the anti-tungsten will be hit by about 19200 molecules per square meter per second. That corresponds to $$1.1551\times 10^{-5}$$ W/m$$^2$$. In terms of heating this can be balanced by keeping the chamber walls somewhat cooler so the antimatter radiates away the heat.

But the effect of the molecular impacts will likely cause trouble. When a proton-antiproton annihilation occurs there are on average 5.3 pions released with average energy 350 MeV, and since they are strongly interacting they can deposit up to 2 GeV in the host nucleus - enough to scatter it (oops, maybe should have used anti-iron instead, not that it would have helped much). Since these are surface nuclei we should expect at least half of the disintegration products to fly off at relativistic speeds (which makes capturing them with the magnetic field hard). We hence end up with on the order of $$10^4$$ antimatter nuclear fragments per second and square meter heading off to cause havoc. Since they are often heavier nuclei they also pack a bigger punch than the initial hydrogen when they hit a wall. Plus, this reaction may dislodge other antimatter atoms.

Can this be fixed? A large chamber area can spread out the damage, but since we are working in a high vacuum there is no way of sucking out the fragments. We can cover the walls with turbomolecular pumps, but they will only act on matter fragments (the antimatter ones will annihilate and release more fragments). The pumps may be able to give them enough momentum to move in some safer direction for disposal but it looks tricky.

Overall, for a sufficiently low pressure (remember, the above value was 100 times above interstellar vacuum) and a large chamber with properly shaped walls I suspect the multiplication rate can be kept low enough that there is just a background loss of a few thousand atoms per second per square meter, enough to keep the antimatter essentially forever. Very clever designs of the magnetic field cradle may help. But it looks like it requires some serious engineering calculations and ingenuity. And any puff of gas accidentally released will quickly start a chain reaction that ends with hot ambiplasma.

This is engineering difficulty/impossibility (not practical as far as we can see, but could perhaps be solved with sufficiently devious tricks and overengineering). One should not confuse it with theoretical impossibility (not allowed by the laws of physics as we know them). A proper theoretical impossibility proof would require showing that e.g. the multiplication rate of fragments is always big, or that there is some in principle reason why antimatter cannot be cooled or made dense enough to crystallise.

• Thank you, this is very helpful, but when you say "Very clever designs of the magnetic field cradle may help" what exactly do you mean? How would this work? – user248541 Dec 17 '19 at 0:53
• Charged particles can be channelled along field lines, and the inhomogeneous (and time varying) field needed to keep the antimatter suspension will push them towards lower field strengths. So the right kind of field may shunt charged fragments towards closed pockets where further annihilation is unlikely to cause more fragments. The design would probably have to be complex. – Anders Sandberg Dec 17 '19 at 1:06
• @HamiHashmi - The magnetic field would be produced by electromagnetic coils outside the walls of the container cylinder. They would need to adjust the field continuously to keep the antimatter centred - a constant field would not be stable (due to Earnshaw's theorem) so one cannot use permanent magnets. (For diamagnetic materials one can levitate them in a stable field, which this paper researchgate.net/profile/Oleg_Semyonov/publication/… uses to store anti-hydrogen). – Anders Sandberg Dec 17 '19 at 12:38
• @PM2Ring - Yes, that introduces a bunch of extra problems (I can imagine magnetically levitating chunks into a reaction chamber through airlocks etc.) Plus, maintenance of the storage chamber will be a minor nightmare. – Anders Sandberg Dec 17 '19 at 18:27
• @HamiHashmi - As I argued in the answer, the magnetic field has to be very strong to keep annihilation-dislodged particles from hitting anything. I think clever shaping of the walls can reduce the annihilation cascades by making antimatter hitting them send matter particles into capture pockets. The shape of the antimatter object is less important, but maybe it can also have fragment-reducing surface features. – Anders Sandberg Apr 15 '20 at 10:31

It is not possible to store this mass of antiprotons , if you see how they can be stored for laboratory uses:

Several hundred antiprotons of 2.1 GeV/c were produced by protons from the PS accelerator and were kept circulating in a machine called ICE (Intial Cooling Experiment) for a period of 85 hours i.e. about 300, 000 seconds (3 × 105). The previous best experimental measurement of antiproton lifetime, acquired during bubble chamber experiments, was about 10-4 second, i.e. a ten-thousandth of a second.

It is an old link, but there are so many orders of magnitude difference to your requirement that I did not bother to look further.

Antihydrogen would seem more promising but is way beyond your needs.

Physicists with the international ALPHA Collaboration at CERN in Geneva have succeeded in storing a total of 309 antihydrogen atoms, some for as long as 1,000 seconds (almost 17 minutes) or even longer -- more than enough time to perform meaningful scientific experiments on confined anti-atoms.

In addition the magnets and Penning traps are such huge instruments , that it would be futile to fit them in a space ship.

So the answer is , no, it is not possible to store antimatter in the large numbers needed for your ideas, even in theory, as far as technology has progressed up to now, and can be projected in the future.