The moiré formed by graphene on Ir(111) was recently shown to be an efficient template for self-organized growth of Ir nanoclusters.¹ The resulting assemblies exhibit extremely high order (less than 12% size distribution in number of atoms per cluster for 70 atoms containing cluster), remain stable well above room temperature, and have an abrupt interface with the graphene support underneath. I am concerned with the generalization of this approach, with the goal of using it to achieve the growth of model, highly ordered magnetic cluster assemblies.

¹ N'Diaye, A., T.; Bleikamp, S.; Feibelman, P.; Michely, Th. Phys. Rev. Lett. 2006, 97, 215501

Thermal catalytic decomposition of carbon containing molecules on a metallic surface, or surface segregation of carbon inside the metal, yields epitaxial graphene on metallic surfaces. We showed that graphene on Iridium exhibit unprecessed structural quality: it preserves its structural coherency across the substrate step edges and through edge dislocations in graphene, i.e. pentagon-heptagon carbon pairs. Depending on the growth conditions, graphene islands with well defined edges (zigzag carbon rows) or full surface coverage of the substrate can be achieved. Graphene on Iridium is a model epitaxial system, with almost unperturated Dirac bands for graphene and clear evidence of a superpotential [moiré of graphene/Ir(111)] upon the electronic structure.

The band gaps of III-nitride semiconductors (GaN, AlN, InN) cover the whole range between ultraviolet and infrared. Owing to a small exciton radius, emission and absorption at well-defined wavelengths can be maintained at room temperature in these semiconductors, when they are prepared as defectless quantum dots (QDs). The growth of nitride nanostructures with controlled optoelectronic, and therefore structural, properties, is still an open question after more than a decade of active research activity. We coupled growth activities with an effort for developping a new technique (EDAFS, MAD) at a synchrotron facility. We accordingly investigated three issues: capping, stacking, and ripening of QDs planes.

A new growth mode during Ir(111) homoepitaxy was recently evidenced thanks to STM investigations.² This growth mode involves decoration rows at twin boundaries, which contaminate Ir homoepitaxy depending on the temperature, and which favor the growth of stacking faults. We compared STM observations to surface x-ray diffraction. Accordingly we quantitatively determined the volumic proportion of stacking faults and showed that volumic faults, which were not yet observed, are formed along <111> directions. We analysed the thermal stability of these defects.

² Bleikamp, S.; Thoma, A.; Polop, C.; Pirug, G.; Linke, U.; Michely, Th. Phys. Rev, Lett. 2006, 96, 115503