Yigal Meir wanted a puzzle to solve during a sabbatical at Princeton University in New Jersey. Meir, a theoretical physicist at Ben Gurion University in Israel, and his host, Ned Wingreen, decided to tackle an anomaly that has baffled condensed-matter physicists for over a decade.

When researchers first studied the way electrons flow through a quantum-point contact, a small constriction connecting large reservoirs, they expected the rate to increase as the gap grew bigger, in steps of a universal value. The rate did, but with an additional first step of about 0.7 times the expected universal value. At first, this '0.7 anomaly' was thought to arise from irregularities in the device. But similar effects were found in almost all experiments involving quantum-point contacts, hinting that something in the quantum gap's environment influences electron flow. This anomaly has been problematic for researchers trying to use quantum mechanics to make smaller, faster computing devices, because such devices demand that no outside environmental factors affect the circuits.

Meir and Wingreen began by studying the available data. They noticed that the anomaly emerged at low temperatures, and that when the temperature was further reduced the conductance began to rise from 0.7 to 1. “This looked very similar to the Kondo effect,” says Meir. “This occurs during the process of electron transport through 'quantum dots' — small spaces to which electrons are confined with the help of several quantum-point contacts.”

Meir and Wingreen asked Charles Marcus, an experimentalist at Harvard University, to look for a signature of the Kondo effect in quantum-point contacts at low temperatures. Marcus confirmed their hunch, but this raised another issue. The Kondo effect describes electron scattering by a magnetic impurity and requires at least one confined electron. Electrons in quantum-point contacts seem to be free from such constraints.

Despite this, Meir and Wingreen decided to invoke the existence of a magnetic impurity in a quantum-point contact to see if that would explain the anomaly. They assumed a magnetic moment in their theoretical calculations that explained the 0.7 anomaly, including its temperature and magnetic-field dependence.

But they needed to identify the proposed impurity to overcome “scepticism from the physics community, which centred on the puzzle of how a magnetic moment could form in such a system”, Meir says. A classical analogy of a quantum-point contact is 'a sea of electrons around a hill'. “The existence of a magnetic impurity on the point contact is equivalent to the formation of a puddle of water at the top of the hill — a counterintuitive phenomenon,” says Meir.

“To find out more, I asked my postdoc, Tomaz Rejec, to develop a spin-density-functional theory calculation that could describe the experimental system,” Meir says. Rejec found that the interplay of spin, electronic repulsion and quantum effects caused a small dip near the top of the 'hill' where an electron and its spin can be captured, forming a magnetic impurity.

This is both good and bad news for quantum computer devices based on quantum dots, says Meir. Magnetic impurities at point contacts would render such computers inoperable. But the magnetic impurity is formed only when conductance through the point contact is around 0.7, so setting the conductance of each contact below that value should allow a circuit formed by quantum dots to function.