Can the neutrons in a nuclear reactor be collimated?

Question

N.B. I am not a physicist.

My layman's understanding of a nuclear reactor is essentially that neutrons are doing one of 4 things at any given time in the reaction chamber:

1. Flying freely around.
2. Colliding with fissile material atoms and causing fission.
3. Colliding with reflectors and bouncing back into the reaction zone.
4. Colliding with control rods and being absorbed.

This process reminds me a lot of the way lasers work, with essentially a mirrored "box" that contains the light and concentrates it before releasing it in a pulse.

So, my question is, could we collimate neutrons in a reactor in the same way we collimate light in a laser? Could we release them in a pulse? Could we essentially make a neutron laser?

Corrolary question: if this is possible, does it then follow that we could safely cooldown a nuclear reactor by "venting" the neutrons into space, by collimating them up and opening an aperture, like a laser?

Answer 1

Speaking as a person who has built a number of neutron collimators: you can not collimate the neutrons inside of a reactor.

Your question seems to be about comparing the motion of neutrons in a reactor to the motion of photons in a laser. Let's talk briefly about how a laser works:

1. You have some "gain medium" which you can undergo a "population inversion," where more of the medium is in an excited state than in the ground state.

2. Some thermodynamically-unimportant fraction of this gain medium relaxes to the ground state, emitting a photon.

3. This photon stimulates the coherent emission of other photons from other parts of the gain medium.

4. The coherent light builds up in intensity because the gain medium is inside of a resonant optical cavity.

5. The optical cavity has such a large quality factor ("$Q$-factor") that the trapped coherent light becomes extremely intense, and some of it is allowed to leak out.

While a long, linear optical cavity is the most common design for macroscopic lasers, like the historically important ruby laser or the various kinds of gas or dye lasers, the optical cavity can actually have any shape. In a diode laser, where the gain medium is a piece of semiconductor, the volume of the optical cavity is defined by the size of the diode, which is extremely small. The laser light emitted by a diode has a huge divergence, and to make a "ray" requires some aligning optics. That's very different from the intrinsically long optical cavity in something like a helium-neon laser, where the light which leaks out is already fairly well collimated, because nonparallel light has a low gain in the cavity.

If you like, you can think of the fissile material in a reactor as a "gain medium" which turns neutrons into more neutrons. But crucially, the new neutrons are incoherent. In laser emission, each stimulated photon has the same energy as the photon which stimulated it. But in a reactor, you need a population of thermal or colder neutrons to induce fission, while the fission neutrons are released in a high-energy state. A crucial part of keeping a reactor going is removing the heat from the new fission neutrons, so that they can thermalize and induce the next fission.

If you are at a reactor and you'd like a linear beam of neutrons, what you do is to cut a hole in the side of the reactor, so that some of the neutrons exit the hole instead of contributing to the reaction. You can keep more of the neutrons if you put a tube between the hole in your reactor wall and the place you'd like the neutrons to go. There is an entire sophisticated technology ecosystem about removing neutrons from a reactor core or other neutron source and bringing the neutrons to a distant experiment.

In such a neutron beam, on the outside of a neutron reactor, a "collimator" is a piece of neutron absorber with a hole in it, to make the beam smaller. For example, you might have a gap in your neutron guide for some apparatus, such as a polarizer or a velocity selector. In the gap, the beam diverges a little bit, and the collimating absorber eats up all of the neutrons who would miss the neutron inlet of your next piece of equipment.

Answer 2

I will answer only simply and quickly, otherwise, it is possible to develop on pages and pages, all this concerning neutron physics.

Yes, it is possible to collimate reactor neutrons and this is very common in research reactors. See "beam tubes", "guide neutrons", "hot and cold sources". It is even possible to sort them in velocity to obtain mono-energetic neutrons.

No, you cannot simply store neutrons: in thermal neutron reactors, their lifetime is about 0.003 s, at the very maximum, inevitably, they are absorbed quickly by the materials of the reactor structures or the fuel when they did not cause fission. They contribute to the neutron activation of materials which become radioactive. An exception: if you have "ultra-cold" neutrons, these bounce off the materials without being absorbed; This is possible in very small quantities if you have a "cold source" and can keep the neutrons in an ultracold state.

Answer 3

As the main question already has answers, let me address the corollary on if we can cool a reactor by venting neutrons.

To keep a nuclear chain reaction going, enough neutrons that are emitted from one fission event have to cause another fission event. That is what we mean by the reactor being critical. Depending on the reactor design, part of keeping a reactor critical may be neutron reflectors placed around the reactor core to reflect neutrons back that would otherwise escape the core. In principle a reactor could be controlled by opening or closing such a neutron reflector. Neutrons that escape would probably not reach space but they would be absorbed by the atmosphere or anything else they would fly toward. But it is much more practical to control a reactor by inserting or extracting control rods rather than opening or closing a big neutron reflector. And also we don't want neutrons flying about, turning anything they hit radioactive.

What you say about cooling down a reactor sounds as if you are thinking about the type of cooling to prevent a meltdown after an accident. But in such cases (e.g. the Fukushima power plant) the reactor would typically already be shut down by fully inserting the control rods, so the reactor core is not critical and (nearly) all neutrons that are still produced will get absorbed instead of causing new fission reactions. But in that situation the reactor core is still producing heat. That heat is by and large not from fission reactions, but because the reactor core is full of unstable atoms (as fission products or activated fuel) that decay, which also produces heat. If that heat is not removed, the reactor core can melt and cause all kinds of big problems. But that heat is not generated from neutron-induced fission but by radioactive decay of unstable atoms. Letting neutrons escape the core won't help, as there aren't any flying around.