The electronic properties of epitaxial graphene are influenced by the structure of the silicon carbide surface that it is grown on. A team of researchers from Sweden and Germany used MAXPEEM as one of the ways to understand the effect. The result points to a promising way of tuning the characteristics of two-dimensional materials on the nanoscale.
Graphene is a two-dimensional material
When you write with your pencil on paper, you get a line made up of graphene – single atomic layers of graphite. Being just one layer of atoms thick, graphene is a two-dimensional material. There are several other 2D materials such as MXenes, boron nitride or molybdenum disulfide. One of the reasons that these materials have become popular is that they have unique properties for advanced electronic devices.
A common way of manufacturing graphene sheets is to grow them epitaxially on top of a foundation of silicon carbide. The results of the present study show that the silicon carbide is much more than just a foundation and can be used to tune graphene properties on the nanoscale.
“Graphene is not just a single carbon layer when it comes to epitaxial graphene on silicon carbide, epigraphene. The epigraphene also includes a so-called buffer layer and the stacking of silicon carbide bilayers under it, where the proximity interaction between the layers is significant,” says corresponding author Davood Momeni at Physikalisch-Technische Bundesanstalt in Braunschweig, Germany.
A way to modulate the properties
Silicon carbide consists of silicon and carbon and can exist in around 250 different forms. The form used as the substrate in this paper is called 6H and has a hexagonal crystal structure.
When producing the silicon carbide wafers, they will always have a slight angle to the atomic planes of the crystal. This angle results in steps and terraces on the surface. Because of the underlying crystal structure, the terraces will have different atomic-scale surface structures, which will affect the properties of the graphene growing on top.
“The graphene properties on the silicon carbide terraces are indeed not uniform, on the contrary to the general assumptions”, adds Momeni.
The researchers produced the graphene investigated in this study using the PASG method, which stands for polymer assisted sublimation growth. With this technique, the standard conventional Si sublimation growth is supported by a polymer which acts as an external carbon-rich source. The polymer adsorbates crack and lead to nucleation at the terrace regions.
During the growth, certain types of terraces-steps will evaporate, or retract, as others stay on the surface. Using the PASG method results in a fast buffer layer formation and retard the step bunching mechanism.
It turns out that the various surface structures of the resulting terraces will give rise to different types of doping and change the electronic properties of the graphene. Doping means that a material will end up with more or fewer electrons than usual. So, by preparing the substrate surface in a certain way, it is possible to tune the characteristics of the 2D material growing on top.
“For the first time, we have shown that there is a doping variation in epitaxial graphene, induced by a proximity effect of the underlying near‐surface silicon carbide stacking order,” says Momeni. “This finding offers a novel growth‐mediated nanoscale doping‐engineering of epitaxial graphene and other stacking materials systems on dielectric polar substrates.”
Investigating the terraces
The research team imaged silicon carbide using several different surface-sensitive methods. They performed low-energy electron microscopy and X-ray photoelectron microscopy measurements at the MAXPEEM beamline, as well as Scanning Tunneling Microscopy, Atomic Force Microscopy and Kelvin Probe Force Microscopy. The result shows that the substrate material will affect the graphene by so-called polarization doping.
“In our study, for the first time, the surface-dependent polarization doping in hexagonal silicon carbide is shown experimentally, which confirms the theoretical predictions, ” says Momeni.
However, there is still a lot to be explored for the graphene on silicon carbide materials system.
“The buffer layer, due to its partial covalent bonds to the silicon carbide, is sometimes considered as being part of the silicon carbide bulk. However, a probable dissimilar binding nature on non-identical silicon carbide terminations is yet remained to unfold,” concludes Momeni. “Moreover, the possible influence of the vacancy/divacancy on non-identical terraces need further investigation. The potential influences of the stacking-order induced doping variation in quantum resistance standard, so far one of the main applications of the epigraphene, still needs to be understood.”