Steven Bramwell (left) and Sean Giblin.

Unlike electricity, which is carried by packets of negative or positive charge in the form of electrons or ions, magnetism was not thought to reside in particle form. It always seemed to exist on a larger scale and to be in 'dipole' form — one end of a magnet is 'north', the other 'south'.

It's the most exciting experiment I've ever done, because it closed the loop between theory and experiment.

But a small cadre of mathematical physicists have been theorizing about magnetic monopoles — discrete atom-sized packets of either north or south forces, analogous to electric charge. Now one team, led by Steven Bramwell at University College London and Sean Giblin at the Rutherford Appleton Laboratory in Chilton, UK, have demonstrated physical evidence for currents of such magnetic charge, which they term magnetricity. “It's the most exciting experiment I've ever done,” Bramwell says, “because it closed the loop between theory and experiment.”

The road to success began with Bramwell's group's 1997 discovery of spin ice — a type of material with unusual magnetic properties1. Spin ice is a frustrated system that has no magnetic order. The pairwise magnetic interactions between spinning atoms in spin ices do not result in ordered arrangements of atomic magnetic dipoles. Instead, these interactions adopt a disordered arrangement analogous to that resulting from unequal bond lengths — another example of frustrated pairwise interactions — in solid water, or ice. “It's a special and interesting arrangement that stays disordered down to absolute zero,” Bramwell says.

In water or water ice, H2O molecules can split to form a small concentration of H+ and OH− ions that drift apart and carry electric current, acting as mobile charges. Bramwell wondered whether the analogous defects in the magnetic structure of spin ice might also drift apart, creating currents. “I spent years and years thinking about how to detect them,” he says.

Inspiration came about two years ago, when a team led by Claudio Castelnovo at the University of Oxford, UK, proposed that magnetic monopoles could emerge in spin ice2. Their theoretical paper suggested that the defects in the spin ice's magnetic structure are effectively magnetic particles that behave like ionic charges. “Only when that paper came out did I start making headway,” Bramwell says. The paper demonstrated, at least in theory, that magnetic defects in spin ice are analogous to the ions in water ice not only in position, “but also in the way they interact”.

This conjecture allowed Bramwell to make a key prediction based on an old equation. Because he has a background in chemistry, he was “vaguely aware” of the equation, put forth by future Nobelist Lars Onsager in 1934, that described a relationship between the strength of an electric field and the disassociation of ions in water3. On the basis of the Castelnovo paper, Bramwell reasoned that a magnetic charge might respond to a magnetic field in the same way that ions in water ice accelerate in response to an electric field. He therefore wanted to apply a magnetic field to spin ice and see whether he could measure the drift of magnetic ions.

The next challenge was how to take such measurements. Fortunately, Bramwell had begun collaborating with Giblin, who works on the muon spectrometer at the Science & Technology Facilities Council's ISIS facility — a 'supermicroscope' that can be used to study materials at the atomic level. “A muon can be used as an atomic bar magnet,” Giblin says. “It is implanted into a sample, where it is incredibly sensitive to local magnetic fields and will give a distinct signal.”

After measuring magnetic fluctuations in spin ice, “It became apparent quite quickly that it was doing what we expected,” Giblin says. They plugged their experimentally derived values into the Onsager equation and the result they obtained — an elementary unit of magnetic charge — precisely matched Castelnovo's 2008 prediction. Giblin's response, he recalls, was “It's real. Let's go to the pub. And then let's measure some more.”

They were so excited, Bramwell says, “We submitted the first draft of the paper to Nature only four days after the end of the experiment.” (See page 956.) With its publication, textbooks may need rewriting. “The result that these things are magnetic charges is quite an extraordinary one — that they move around freely of each other,” Bramwell says. “It doesn't break the laws of physics, but magnetic charge has not previously been thought to be something that can carry current like electric charge at the atomic scale.”