Gigantic Antarctic Instrument, IceCube, Finds Mysterious Cosmic Neutrinos

A cubic kilometer-sized neutrino detector buried deep in Antarctic ice has confirmed the discovery of mysterious ultra-high energy neutrinos

7 min read

Evan Ackerman is IEEE Spectrum’s robotics editor.

Gigantic Antarctic Instrument, IceCube, Finds Mysterious Cosmic Neutrinos
Photo-illustration: IceCube Collaboration

In a research published last week in Physical Review Letters, an Antarctic detector has managed to confirm the existence of a small handful of cosmic neutrinos: ultra high energy particles that likely originated from unknown sources far outside of our galaxy. These particles are nearly impossible to detect—it took a specially designed system of sensors burried in a cubic kilometer of ice. But the few we've seen have energies up to a thousand times greater than what the Large Hadron Collider can generate, and we're just starting to be able to get a sense of where they might be coming from.

(Go here for the incredible story of the engineering of the detector, called IceCube.)

What are Neutrinos?

Neutrinos are elementary particles (like quarks, photons, and the Higgs boson) that have no charge and virtually no mass. Since they're small, fast, and charge-free, they aren't affected by nuisances like electromagnetic fields, meaning that they can pass unmolested through rather a lot of pretty much anything. Some 65 billion solar neutrinos just passed through every square centimeter of your body, and if you wait a second, 65 billion more of them will do it again. The only way to bring a neutrino to a halt is if it runs smack into an electron or the nucleus of an atom, but this is ridiculously improbable: you'd need a piece of lead about a light year long to be reasonably sure catching any one specific neutrino. Fortunately, the enormous number of neutrinos that are flying through everything all the time compensates for the low probability of collision, and that has allowed us to learn some things about these elusive particles.

We’re pretty sure that neutrinos come in three different (but equally tasty) flavors: electron, muon, and tau. Each flavor has a slightly different (but tiny) mass, on the order of a million times smaller than the mass of a single electron, and a neutrino can oscillate between these three flavors as it zips along. The original flavor that each neutrino takes depends on how it was created: most often, neutrinos are created through high energy nuclear processes like you'd find going on inside stars. To take one common example, protons and neutrons colliding with each other create pions, which are subatomic particles that decay into a mix of muon and electron neutrinos. 

The most common source for the neutrinos that we see here on Earth is the sun, which produces an electron neutrino every time two protons fuse into deuterium. (This happens a lot.) What's much rarer to see are neutrinos that aren't produced close to home—neutrinos that come from outside of our solar system, and even outside of our galaxy. These are called cosmic neutrinos, or astrophysical neutrinos.

Cosmic Neutrinos

Cosmic neutrinos are born, we think, in the same sorts of ultra high energy events out in the universe that also generate gamma rays and cosmic rays. We're talking events like supernova remnant shocks, active galactic nuclei jets, and gamma-ray bursts, which can emit as much energy in a few seconds as our sun does over ten billion years. As you might expect, the resulting neutrinos have stupendously high energies themselves: a million billion electronvolts (1 petaelectronvolt) or so. That works out to be about a tenth of a millijoule, which is a lot for a particle that has an effective size of nearly zero, and it is about equivalent to the kinetic energy of a thousand flying mosquitoes, in case that horrific unit of measurement is of any help to you.

The reason that cosmic neutrinos are important is the same reason that neutrinos themselves are so frustrating to measure: they ignore almost everything. This allows them to carry valuable information across the entire universe. What sort of information can be carried by a particle that we can't even measure directly? Here are three examples:

  • Since neutrinos aren't affected all that much by even the densest matter, they can escape from the core of a supernova well before the shock wave from the inside of the collapsing star makes it to the outside and releases any photons. By detecting this initial neutrino burst, we can get a warning of as long as a few hours before a supernova would be visible from Earth.
  • Since neutrinos aren't affected at all by magnetic fields and therefore travel in straight lines, they can help us pinpoint the origin of ultra high energy cosmic rays, which are affected by magnetic fields and therefore can follow winding paths. We know that some cosmic rays come from supernovae, but many of them don't, and we're not sure where the rest of them originate. With energies over a million times greater than the Large Hadron Collider, it would be nice to know where they come from.
  • The ratio between different flavors of neutrinos may suggest how they were formed. For neutrinos produced by pion decay, we'd expect to see two electron neutrinos for every muon neutrino. If we see different ratios, it would suggest a different formation environment, and particularly weird ratios could even lead to new physics.

Detecting Neutrinos

Since neutrinos don't really interact with anything, there's no reliable way to detect them directly. Whether they're coming from the sun, the atmosphere, or somewhere more exotic, the best we can do is try and spot the aftermath of a very unlucky neutrino smashing headlong into something while we watch. 

When a neutrino does smash into something, one of two different things can happen. If the neutrino isn't super energetic, it might just bounce off in a new direction, passing on some of its momentum and energy into whatever it hits (which recoils in response) and occasionally causing that thing to break into pieces. Some neutrino detectors are designed to watch for both of these effects. The other thing that can happen is that the neutrino obliterates itself, dissolving into a subatomic particle that depends on the neutrino's flavor-of-the-moment: an electron neutrino turns into an electron, a muon neutrino turns into a muon, and a tau neutrino turns into a tau particle, stripping electric charge off of whatever it hits as it does so. Some detectors look for this change in charge of the thing that the neutrino ran into (it's the only way of detecting neutrinos with energies less than 1 MeV), but the detector that we're interested in looks for traces of the ex-neutrino's subatomic particle itself.

Detecting particles moving at very high speeds is a (relatively) straightforward thing, at least conceptually. The key is what's called Cherenkov radiation, which is created by charged particles moving through a medium faster than the speed of light in that medium. It's sort of like the sonic boom created by an object travelling through air faster than the speed of sound, except with with light. Here's what the Cherenkov radiation created by a burst of beta particles from a nuclear reactor being pulsed looks like:

So now that we've got some lovely blue glowy-ness to look for, all we need a medium through which charged particles can travel faster than light, and some kind of detection system that can spot the resulting Cherenkov radiation.


The IceCube Neutrino Observatory is a massive neutrino detector in Antarctica that takes advantage of the fact that the south pole is covered in a medium through which charged particles can travel faster than light: ice. A kilometer down, the ice is beautifully clear, allowing bursts of Cherenkov radiation to propagate through it unhindered. The IceCube observatory itself consists of 5,160 digital optical modules, each about 25 centimeters in diameter, suspended on 86 individual strings lowered into boreholes in the ice:

Each string sits between 1450m and 2450m below the surface, spaced 125m horizontally from neighboring strings, resulting in a neutrino detector that's a full cubic kilometer in size.

imgIllustration: IceCube Collaboration

What IceCube is looking for are the tiny flashes of blue light emitted by the electrons, muons, and tau particles flashing through the ice after a neutrino collides with a water molecule. These flashes are very dim, but there are no other sources of light that far down under the ice, and the photomultiplier tube inside each digital optical module can detect even just a handful of photons

Depending on what kind of subatomic particle the neutrino turns into, IceCube will detect different Cherenkov radiation patterns. An electron neutrino will produce an electromagnetic shower (or cascade) of particles that looks like this:

The muon produced by a muon neutrino, on the other hand, can travel hundreds of meters, leaving a track that points back along the same trajectory as the muon neutrino that created it:

A tau neutrino will produce a sort of combination of these two signatures. Maybe. We think. Tau neutrinos are difficult to detect, because tau particles themselves are extraordinarily massive and short lived: they're something like 3,500 times the mass of an electron (and 17 times the mass of a muon), with a lifetime of just 0.0000000000003 second, which means that they decay into other subatomic particles virtually instantaneously and are easily mistaken for electron neutrinos. IceCube has some ideas of what unique radiation signatures might suggest the detection of a tau (including the "double bang," the "inverted lollipop," and the "sugardaddy") , but they haven't found one yet.

Muons Everywhere

IceCube has been focused on measuring muon neutrinos, because the muon track shows exactly where the neutrino came from to within about a degree, which is a valuable bit of information to have. And IceCube detects plenty of muons, because they're produced constantly by cosmic rays smashing into molecules in the upper atmosphere. In order to weed out the cosmic ray muons and focus on just the muons produced by muon neutrinos, IceCube has been measuring only neutrinos coming from the northern hemisphere. Situating a northern hemisphere observatory at the south pole seems a little weird, but this way, IceCube can use the entire Earth as a filter to keep out those pesky atmospheric muons. By only tracking muons that arrive from "down," coming up through the Earth itself, IceCube can be sure that if it sees a muon, it must have been generated by an actual muon neutrino, and nothing else.

Latest Results

IceCube detects tens of thousands of neutrinos (and billions of muons) every year: about one every six minutes. But, only a very small number of those detections represent neutrinos that have high enough energies and the right trajectory that they can confidently be described as cosmic neutrinos. Based on nearly two years of data acquired between 2010 and 2012, IceCube has identified perhaps a dozen neutrinos that almost certainly have cosmic origins. Here's a map of where those neutrinos (the red circles) came from; the green circles include earlier data as well as electron neutrinos that don't offer much in the way of trajectory information:

imgMap: IceCube Collaboration

What's notable about this map is that it doesn't show any clustering: the muon neutrinos didn't all come from the same place, and there aren't even isolated clumps of them coming from a similar spot. Specifically, there's no correlation with the plane of our own Milky Way, suggesting that these cosmic neutrinos came from outside of our galaxy, and that there must be a whole bunch of extragalactic sources producing these things.

There's no reason to be disappointed by these results: if anything, it reinforces the fact that our universe is that much more vast and strange. And remember, we've only just started looking for cosmic neutrinos within the last half decade, and while gobs of them have probably passed through you while you've been reading this article, a square kilometer sized detector in Antarctica has spent two years looking and only managed to spot a dozen or so. There's a lot more to learn, and a lot more impending discovery: the data that this paper is based on are three years old, and IceCube's detectors are still peering deep into the ice at the south pole.

The Conversation (0)