The emphasis of this project has been on the investigation of and the search for topological materials, in which magnetism or strong correlations play an important role. To this end, we have employed and further developed novel analysis tools based on resonant x-ray absorption (XAS) and reflectivity (RXR), combined with crystal and ligand field theory. These tools allowed us to measure parameters such as spin-orbit interaction energies, correlation strengths and magnetic moments, which control the delicate phase interplay in correlated and topological materials. We have validated our analysis tools by investigating the bulk properties of SrTiO3 , and by uncovering surface orbital reconstruction at the surface of TiO2 terminated SrTiO3 . We have also revealed electronic reconstruction at the LaCoO3 surface by studying the Co valence profile.
Then, we have studied Mx (Bi,Sb)2−xTe3 magnetically doped with M = Cr, V and Eu using a com- bination of XAS/XMCD with ligand field theory calculations. Our results invalidate the experimental foundation of previous claims of van Vleck magnetism as the main mechanism of ferromagnetism in V-doped (Bi,Sb)2Te3 . Our multiplet calculations for Cr and V and x = 0.1 confirm the substitution of the magnetic impurity atoms into the lattice at cation sites rather than into the van der Waals gap, and we observe no indication of impurity clustering. We further determine the configuration composition of the ground state, the magnetic moments and further parameters such as the crystal field splitting. Overall, our results demonstrate the importance of p-d hybridization for the magnetism.
In the course of the further work on this project, we have identified strongly correlated 4f systems as very promising: They have stronger spin-orbit coupling than 3d elements, and the ideas for the realization of topological physics are further advanced than in 3d-metal oxides. Also, their frequent mixed-valence character, large moments and correlations can be well studied by resonant x-ray tech- niques. Hence, in collaboration with C06 we have studied SmB6 , a mixed-valence, strongly correlated Kondo insulator and a promising contender in the quest for one of the first systems with topologically non-trivial and correlated surface states. Using RXR combined with crystal field calculations, we have uncovered chemical and valence surface reconstructions, and have studied their dynamics. Our re- sults indicate a boron termination and subsurface dominance of Sm3+ over Sm2+ . This reconciles earlier studies that were based purely on surface sensitive techniques such as ARPES and STS.
With the very successful work on SmB6, the emphasis of the project has shifted towards such 4f systems, and we have decided to finalize C04 and to join forces with C06 in the second funding period.