Here’s a question you might not have asked yourself before. When protons collide in the Large Hadron Collider, how often are undetectable particles produced?

Maybe a quick follow up question would be, why does it matter?

A paper I have been working on for several years has just been accepted by the journal for publication, which is nice news to receive while on holiday, and it answers the first question, at least under certain conditions. Louie Corpe already wrote about the paper when the ATLAS collaboration (of which we are both members) first released it and submitted it for publication. Now it is accepted, we will be releasing the data (to HEPdata) and relevant code (to Rivet) as soon as possible, so it can be reused, hopefully widely and for a long time, which was the main point.

Why it matters.

From various astrophysical observations, it seems likely that around 80% of the mass of the universe is undetectable, in the sense that, while we see the gravitational influence of enormous clumps of it, if it is made of particles they interact so weakly with normal matter that all our detectors are blind to them. Astronomers and Physicists refer to this as “Dark Matter”, although as has been pointed out before, “Transparent Matter” would have been a more accurate, if less evocative, name.

It might be that whatever this Dark Matter is, it can be produced in collisions at the Large Hadron Collider (LHC) at CERN. For what it is worth, many theories predict this. If you care what 80% of the universe is made of (and why wouldn’t you?) then this matters.

Not a null result

If Dark Matter were to be produced in a collision, it would escape our detectors without registering directly. Since our detectors surround the collision point¹, however, we have a chance of inferring its presence from the fact that we would see an imbalance when we add up the momentum of all the other particles produced in the collision.

Searches for Dark Matter have been done before at the LHC using this effect, and have found no sign of it. Spoiler alert, we also find no sign. This analysis is different from the previous searches though, in what I strongly believe is a crucial respect. It is more than just a null result.

Now, bear with me. I know null results are important, and we are privileged to have a culture in particle physics where we can readily publish them. They add to the sum of knowledge by ruling out possibilities, and they avoid future wasted effort.

I say that our paper is not a null result though because undetectable particles are produced at the LHC, and we have measured how often that happens. The particles in question are called neutrinos, tiny particles which are too fast moving to explain the oberved effects of Dark Matter, but which are an essential component of the “Standard Model” of particle physics.

The measured production rate agrees quite well with the best predictions of the Standard Model, although in detail there are some discrepancies, which we have quantified. The discrepancies don’t look as though they are due to Dark Matter though, and may well be to do with deficiencies in the predictions. Time will tell.

Anyway, measuring a physical process at higher energies and more precisely than ever before, and using that to test our theory in a new regime, is not a null result by any stretch.

Bonus

And there is an added bonus in that now this measurement has been made, it can be used in lieu of dedicated searches for Dark Matter, to rule out possible new theories – something we demonstrate in the paper with a couple of examples. Null results are important, but when theories proliferate it is important to be efficient when confronting them. A dedicated search for each of them might be optimal for each theory, but is not efficient. A single measurement of what really does happen, made in a theory-neutral way, can do a good enough job for many possibilities, now and far into the future.

¹apart from the where the proton beams come in and go out. So strictly we see an imbalance in the momentum transverse to the beams.