The magnetic properties of cobalt atoms lying on the surface of graphene can be controlled by the choice of substrate under the graphene sheet. This unexpected discovery has been made by physicists in Switzerland and could be exploited someday to create extremely dense magnetic memories or even quantum bits (qubits) for quantum-information processing and storage.
Graphene is a sheet of carbon just one atom thick. It has a number of unique electronic and mechanical properties that could be used to create new types of electronic technologies. These include spintronics, which aims to make use of the spin magnetic moment of the electron in circuits that are smaller, faster and more energy efficient than conventional electronics.
Now, Harald Brune and colleagues at the Ecole Polytechnique Fédérale de Lausanne (EPFL) and the Swiss Light Source (SLS) in Villigen have discovered an effect that could be exploited to create spintronics devices based on graphene. “The magnetic properties of transition-metal atoms on graphene were, so far, thought to depend only on the transition metals themselves,” explains Brune. “However, in almost all experiments, we need a substrate on which to grow graphene, and in our new work we show that this substrate greatly influences the magnetic properties of the transition metals that find themselves on top of it.”
In previous work, Brune and colleagues placed atoms of the transition metal cobalt on a graphene surface that had been grown on a platinum substrate. They found that the cobalt atoms have a magnetization that is in-plane – that is, pointing parallel to the surface of the graphene. However, in this latest work they discovered that when the graphene is grown on a ruthenium substrate, the magnetic moment of cobalt points out-of-plane. They also tried an iridium substrate and found that like platinum, the cobalt moment lies in-plane.
“The substrate thus plays a much more important role than previously thought and calculations, which until now considered graphene as freestanding, need to take this into account,” he says. “Our result also shows that we can actually tailor the magnetic properties of the transition-metal atoms, depending on the substrate they lie on.”
The graphene films were grown on ruthenium and iridium substrates using chemical vapour deposition. The graphene was then given a sparse coating of cobalt atoms using electron-beam evaporation.
The researchers made their measurements on samples at 3.5 K using X-ray absorption spectroscopy (XAS) and X-ray magnetic circular dichroism (XMCD) techniques. XMCD measures the magnetic properties of an atom using circularly polarized X-rays. “If the X-rays are polarized, we can infer whether the magnetic moments of the cobalt atoms lie along the direction of the incoming X-rays or in a direction perpendicular to them, and we can also calculate the size of this magnetic moment,” says Brune. “Applying an external magnetic field allows us to determine how much field is needed to align the magnetic moments of the individual cobalt atoms being probed.”
The team found that the magnetic properties of the cobalt are influenced by the strength of the bond between the carbon atoms in graphene and the substrate atoms. There are strong chemical bonds between the carbon atoms and ruthenium, for example, whereas there are much weaker weak van-der-Waals interactions with iridium and platinum substrates. As a result, the graphene is pulled much closer to the ruthenium substrate than it is to the platinum or iridium. The distance between the graphene and the substrate affects the graphene, which in turn affects the cobalt atoms.
“Put simply, we can imagine that the underlying metal surface transfers part of its electrons to the graphene, or the other way around, and this influences the electronic properties of graphene. In turn, this influences the magnetic properties of the cobalt atoms.”
If the magnetic states of transition-metal atoms on graphene are found to endure for long times, they could be used to create extremely dense information-storage devices. They could even be used as qubits, although Brune points out that they would have to be operated at extremely low temperatures.
The team says that it is now focusing its attention on identifying single atoms or molecules that have sufficiently long-lasting magnetic states, so that such applications might indeed be possible one day. “Ultimately, we might be able to store one bit of information in the magnetic state of a single transition metal atom,” says Brune. “Currently, magnetic hard disks use 107 atoms per bit.”
The research is described in Physical Review Letters.
- This article first appeared on nanotechweb.org