Brazilian researchers investigate surface physics at nanometric scale

Application horizon includes more efficient catalysts and devices capable of storing more data per unit area.

By José Tadeu Arantes

Agência FAPESP – A great many technological applications with growing relevance to daily life are driven by phenomena that involve surface materials, ranging from prosthesis-bone interfaces in medicine to automotive and industrial catalysts to data storage and processing devices.

A significant contribution to Brazilian research in the field of material science has come from the Thematic Project “Electronic and geometric structure of nanomaterials: synchrotron radiation studies”, supported by FAPESP.

“All of the phenomena due to interactions between a surface and the external environment, such as adhesion, lubrication, corrosion, and catalysis, basically depend on electron exchanges between neighboring atoms. So, when studying the interaction of a solid with the environment, we need to know what atoms the solid’s surface is made of, how those atoms are distributed on its surface, and how they bond chemically,” said physicist Richard Landers, a professor at the University of Campinas (UNICAMP), São Paulo State, Brazil, and principal investigator for the project.

“The number of surface atoms, about 10¹⁵ per square centimeter, is very small compared with the number of inner or bulk atoms, which is about 10²³ per cm³,” he said. “It’s these 10¹⁵ atoms per cm² that determine how a solid interacts with the environment.”

UNICAMP’s Laboratory of Surface Physics (GFS), led by Landers and Professor Abner de Siervo, pursues four main lines of research in surface science: geometric and electronic structure, morphology, magnetic properties, and solid-gas interactions.

“We don’t aim for immediate applications in our research. However, one of the projects we’re working on could be used to develop more effective magnetic readers,” Siervo said.

“It was the PhD research of Luís Henrique de Lima, a student member of the group. The technique consisted of growing a graphene sheet on a silicon carbide surface and then obtaining clusters of a few atoms of magnetic material on the graphene. In our case, this material was cobalt.”

The graphene was obtained by heating silicon carbide (SiC) to over 1100°C. At high temperatures, the chemical bonds between carbon and silicon break down, the silicon evaporates, and the remaining carbon reorganizes to form a graphene sheet. The resultant sheet is a crystalline structure that is only one atom thick, in which the carbon atoms form hexagonal meshes.

“In 3D terms, the surface obtained on this substrate was shaped like an egg carton, with bumps and dimples in a lattice arrangement. The idea was to distribute the magnetic particles of cobalt in the dimples or cavities, which is where the eggs would be in the analogy,” Siervo said.

“Using scanning tunneling microscopy and X-ray photoelectron spectroscopy, we observed that when the nanoparticles of cobalt on top of the graphene were heated, they migrated below the surface, where they intercalated with the graphene, which protected them from the outside. The key point,” he explained, “is that these particles appear to occupy positions just like eggs in a carton. In other words, we created an organized structure of ferromagnetic nanoparticles separated by precise, known distances that were protected by the graphene sheet.”

Figure 1 – (a), (b), (c) Scanning tunneling microscope (STM) images of graphene grown on silicon carbide (SiC) substrate; (d) representation of cobalt (Co) nanoparticles grown on graphene/SiC; (e) STM image of intercalated Co clusters protected by a graphene sheet; (f) atomic model of intercalation.

One direction in which the research group could develop these findings involves growing an antiferromagnetic oxide on graphene to produce ferromagnetic and antiferromagnetic structures separated by a single layer of carbon atoms.

“The tiny gap between these two structures would trigger an exchange bias, which is the principle by which magnetic readers operate. The difference in this case is that instead of operating at the microscopic scale, our device would operate at the nanometric scale, potentially affording many additional advantages,” Siervo said. “However, before we think about possible applications, we need to understand the physical principles of a regime that hasn’t yet been studied in great depth, which is when nanoparticles are closely packed by being concentrated in large numbers per unit area.”

Model catalysts

Catalysts are a particularly interesting research focus because of their many potential applications. To describe a catalyst exhaustively, it is necessary to know which atoms it is made up of, their positions on the surface, and how they interact (as oxides, carbides or metals, for example). All of these factors influence a material’s interaction with the external environment and therefore govern how it performs its function as a catalyst.

“It’s very hard to obtain all of this information,” Landers said. “To analyze the crystallographic and electronic structure of a surface, you need a method that’s highly sensitive to the 10¹⁵ atoms per cm² that make up the surface layer. X-rays aren’t suitable because they penetrate thousands of angstroms into the substrate and ‘see’ the atoms below the surface. We therefore opted for a specific band of synchrotron radiation, using equipment at the National Synchrotron Light Laboratory (LNLS) in Campinas.”

The photons produced by synchrotron radiation tear electrons away from some of the surface atoms, and the characteristics of the surface can be determined by measuring electron interference during their transit.

“Photon energy can be controlled with great precision by varying the parameters of the light source, so it’s also possible to define which electronic layer of the surface atoms will be excited. The lower the kinetic energy of the photons, the further away from the atomic nuclei will be the detached electrons. This tells you the thickness of the layer you’re analyzing,” Landers said.

From the energy of the detached electrons, it is possible to determine the nature and chemical states of the atoms. Furthermore, if the surface has an orderly crystal structure, the positions of the atoms can be determined based on the directions taken by the electrons. “That’s why the entire project was designed with the use of synchrotron light in mind,” Landers noted.

The researchers at GFS are not only interested in studying surfaces but also in creating them. Their laboratory has an ultra high-vacuum chamber for this purpose. “Under normal atmospheric pressure, in the range of 1,000 mbar, each atomic site on the surface is hit about 1 billion times per second by molecules from the external environment. If we work with a high vacuum, in the range of 10^-9 mbar, the number of hits will fall to 1 per second. But that’s not sufficient because even in a single second the surface will be contaminated. We need an even more extreme vacuum, an ultra-high vacuum, in the range of 10^-10 or 10^-11 mbar,” Landers said. “In our experiments, we’re able to grow monocrystalline layers under controlled conditions using our ultra-high vacuum chamber as part of the analysis system.”

 

Figure 2 – Model catalyst comprising rhodium (Rh) nanoparticles grown on orderly ultrathin magnetite film supported on palladium (Pd). Left and right: photoelectron diffraction patterns with chemical element selectivity.

Real catalysts consist of either an oxide or zeolite (a porous amorphous material) onto which active metals are deposited. Studying these catalysts on an atomic scale is extremely difficult; therefore, the researchers at GFS opted to create and study model catalysts.

“We created a sufficiently thin oxide surface to mimic actual oxide. We used this as a substrate on which to grow particles of metals such as rhodium, platinum or palladium, and investigated the structure formed,” Landers said.

“A substantial proportion of the project funding was spent on the infrastructure required by this investigation. For example, an analysis system such as this weighs more or less 1,000 kg. In addition, a computational cluster had to be implemented to process the experimental data.”

The researchers also required a scanning tunneling microscope because the electron diffraction technique used for surface mapping cannot determine the actual size of the metal particles created.

“The scanning tunneling microscope provides an image of the surface that allows you to distinguish individual atoms,” Siervo explained.

“Most of the time, you can see the nanoparticles obtained and identify the arrangement of the atoms in each one. By combining these two techniques, spectroscopy and microscopy, we were able to achieve a significant enhancement of our knowledge on the systems we were investigating.”