RuO₂ Revisited: Nonmagnetic Bulk, Ferrimagnetic Surface, and a Rutile Roadmap to Spin Control

Altermagnets and correlated oxides promise dissipationless control of spin, yet the most interesting magnetic textures may live where we least expect them: at surfaces and in long-range exchange paths. In this talk we will focus on our recent re-examination of rutile RuO₂, long discussed as a prototypical metallic altermagnet. State-of-the-art DFT, chemical bonding analysis, and simulated STM/ARPES show that bulk RuO₂ is in fact nonmagnetic, but its (110) surface undergoes a spontaneous electronic and magnetic reconstruction. Reduced Ru-O coordination and local symmetry breaking drive a Stoner-like instability that stabilizes robust ferrimagnetic order confined to the topmost Ru-O layers. The resulting spin-polarized surface states exhibit a sizable exchange splitting of 0.5-0.6 eV near the Fermi level, directly accessible by spin-resolved ARPES and tunneling spectroscopy. I will discuss how this “hidden” surface magnetism reconciles conflicting experimental reports on RuO₂, and what it implies for spin-orbit torques, anomalous Hall signals, and catalysis in RuO₂-based heterostructures. To place these results in a broader context, I will briefly highlight our companion work on rutile transition-metal difluorides, where an unusual super-superexchange path TM–F···F–TM resonates with covalency to generate giant, chiral magnon band splittings comparable to nearest-neighbor exchange. Together, the RuO₂ surface and rutile TM-fluoride magnonics case studies illustrate a common theme: by engineering orbital bonding, exchange pathways, and symmetry breaking in simple rutile lattices, one can create large, tunable spin polarizations without net bulk magnetization, opening new routes to interfacial spintronics and THz magnon devices.