Magnetism in thin layers: One electron makes the difference

Graphene was a milestone. When Andre Geim and Konstantin Novoselov produced the first single-atom thick layer of carbon in 2004, they had no idea they were establishing a completely new field of research. Previously, though, ultra-thin layers just a few atoms thick – so-called two-dimensional crystalline materials – had already revealed astonishing optical, electronic, magnetic, and even superconducting properties. In this area, Thorsten Schmitt’s team is among the top research leaders internationally.

The researchers in the Spectroscopy of Quantum Materials group at the PSI Center for Photon Science specialise in the production and spectroscopic investigation of thin atomic layers made up of different chemical compounds, which they stack on top of each other alternately – like a sandwich. They keep finding interesting phenomena in these hybrid materials – most recently in a superlattice in which layers of lanthanum nickelate (LaNiO3) and lanthanum titanate (LaTiO3) alternate. Lanthanum nickelate is non-magnetic (paramagnetic), and lanthanum titanate is antiferromagnetic (see box below: «Magnetism briefly explained»). When the two materials are stacked on top of each other, electrons – negatively charged elementary particles – jump from the titanate to the nickelate, and thus the magnetism changes: Lanthanum nickelate becomes antiferromagnetic, while lanthanum titanate is now non-magnetic.

A sensation in physics

To someone outside the field, this exchange may sound like a useless sleight-of-hand trick. For researchers in physics, however, it is recognised as an important step, because in this way they can tailor materials to enable novel applications, for example, for magnetic memories of the future. Lanthanum titanate is not well suited for this purpose, since as an insulator, it does not conduct electrical current. Lanthanum nickelate, on the other hand, is a good conductor and, in combination with the new magnetic property, a promising starting material for the building blocks of so-called spintronic computers. In these, the classical ferromagnetism of a computer hard drive is replaced by antiferromagnetic memory cells based on the spins of electrons. For such components, the properties on the surface or at interfaces are crucial. «Our research is not oriented towards developing such memories ourselves, but rather towards understanding fundamental properties that define the functional principle for future applications – we are doing basic research,» Schmitt emphasises. This is urgently needed, because many phenomena in two-dimensional materials are not yet understood, and researchers are constantly making surprising discoveries.

The team at PSI has a high level of expertise in both special X-ray spectroscopic examination methods and the production of thin layers of so-called quantum materials. First, to produce the sample: Take two pellets with the two starting materials. These pellets contain the substance in polycrystalline form; however, single-crystal materials are required for special research purposes and future applications. To obtain single-crystal films of the desired composition, PSI researcher Milan Radovic uses a laser to shoot material out of the pellet, which is then deposited on a crystalline substrate. The crystal structure of the material aligns itself with the lattice of the substrate, so the researchers get exactly the crystal lattice they want. Then the same procedure follows with the pellet of the second material, then again with the first, and so on – until a total of 60 layers are stacked on top of each other in perfect alignment. This takes several hours, with the layer thickness being checked after each step. In this particular experiment, the titanate layers were two lattice cells thick, the nickelate layers up to ten lattice cells. The thinnest layers measure less than one nanometre.

«In the past, other research groups have tried this too, but they haven’t managed to do it so perfectly,» says Radovic. That’s because lanthanum nickelate and lanthanum titanate require different background pressures of oxygen to grow in the desired composition. Teguh Asmara and Milan Radovic had the idea of ​​using nitrogen as a background gas to ideally set identical background pressures for both materials.

The tricky part

Then came an even trickier part of the project: Did the change they were hoping to see in the magnetic properties actually occur? To find the answer to this question, the researchers needed to proceed like detectives, in three steps. 

First Andreas Suter, from the Laboratory for Muon Spin Spectroscopy at the PSI Center for Neutron and Muon Sciences, bombarded the sample with muons from the PSI muon source, with the energy adjusted so precisely that these particles stop near the interfaces in the lattice. In an external magnetic field, these muons perform a wobbling gyroscopic motion that can be measured. If this precession remains the same over time, the material is not magnetic at this point. If it decreases over time, this is a sign that the material is magnetic – as in this case. The muons then decay after a short time, emitting positrons that can be used to indirectly detect the behaviour of the muons.

However, the physicists did not yet know what type of magnetism is present. To find out, they measured the sample in a second step with highly sensitive magnetic sensors – without a measurement signal. Since these sensors can only detect ferromagnetism, it was clear that the sample is magnetic, but not ferromagnetic, so it must be antiferromagnetic.

For this, the electron structure is crucial. To characterise that, in a third step, the physicists employed the RIXS method (resonant inelastic X-ray scattering). Using X-rays from the Swiss Light Source SLS with precisely adjusted energy, an electron is lifted from the lowest energy band of an atom into an unoccupied energy band. Because this leads to an imbalance in the atom, another electron from a lower energy band falls into the gap below in a quadrillionth of a second – it also emits X-ray light, but with lower energy. The energy difference between the incident and emitted light provides information about the gaps between the energy bands. The electrons also perform gyratory movements that spread through the material as spin waves, so-called magnons – like a stone thrown into water. In this way, the researchers can probe the complex electronic structure and local magnetic properties of the material.

All possibilities in one place

It is a major advantage that, at PSI, all the requisite research methods are available in one place, Thorsten Schmitt emphasises. His team specialises in spectroscopic studies using X-ray light from SLS. «The colleagues who work with muons sit in the neighbouring office. So we can exploit synergies to the maximum.» Nevertheless, the facilities at PSI are not always sufficient to answer all research questions. SLS is getting on in years, and detailed measurements require other synchrotron radiation sources such as the ESRF in Grenoble or the Diamond Light Source in Oxford. The samples for the current paper were therefore first examined in detail at SLS at PSI to determine the electron structure. Then the two facilities mentioned in France and the UK were used for targeted measurements of the spin waves with the highest possible resolution. 

In the future, through the SLS 2.0 upgrade currently in progress, the facility will deliver more X-ray light and enable more precise measurements in a shorter time. To be able to take advantage of that, Thorsten Schmitt’s research group needs to rebuild the RIXS facility. Instead of the previous five metres, the new facility will be 11 metres long. Also, all of the diffraction gratings for analysis of the X-ray beam will be modernised and configured for greater measurement accuracy. The concept stands ready, the funding is approved, and preparations for the purchase and fabrication of new components are currently under way.

Thorsten Schmitt already has ideas for further research. Researchers recently discovered that two-dimensional nickelates become superconducting when they are placed under high pressure. Because the hybrid material from the PSI project has the same electron structure, it too might become superconducting under similar conditions – that is, conduct electrical current without loss. To achieve this, researchers could place the material in a pressure chamber or use mechanical strain to contract the lattice spacing, thus possibly inducing superconductivity. Doping the material – deliberately contaminating it with foreign atoms – could also be an option. This principle is used for materials that become superconducting at high temperatures. «We absolutely have to try to make our hybrid material superconducting – the topic is simply too promising to put aside,» Schmitt concludes.