Graphene in spintronics has so far primarily meant spin current leads of high performance because the intrinsic spin-orbit coupling of its pi-electrons is very weak. If a large spin-orbit coupling could be created by a proximity effect, the material
could also form active elements of a spintronic device such as the Das-Datta spin field-effect transistor, however, metal interfaces often compromise the band dispersion of massless Dirac fermions. Our measurements show that Au intercalation at the graphene-Ni interface creates a giant spin-orbit splitting (~100 meV) in the graphene Dirac cone up to the Fermi energy. Photoelectron spectroscopy reveals hybridization with Au-5d states as the source for the giant spin-orbit splitting. An ab initio model of the system shows a Rashba-split dispersion with the analytically predicted gapless band topology around the Dirac point of graphene and indicates that a sharp graphene-Au interface at equilibrium distance will account for only ~10 meV spin-orbit splitting. The ab initio calculations suggest an enhancement due to Au atoms that get closer to the graphene and do not violate the sublattice symmetry.
Spin currents which allow for a dissipationless transport of information can be generated by electric fields in semiconductor heterostructures in the presence of a Rashba-type spin-orbit coupling. The largest Rashba effects occur for electronic surfa
ce states of metals but these cannot exist but under ultrahigh vacuum conditions. Here, we reveal a giant Rashba effect ({alpha}_R ~ 1.5E-10 eVm) on a surface state of Ir(111). We demonstrate that its spin splitting and spin polarization remain unaffected when Ir is covered with graphene. The graphene protection is, in turn, sufficient for the spin-split surface state to survive in ambient atmosphere. We discuss this result along with evidences for a topological protection of the surface state.
