Intercalation of Au can produce giant Rashba-type spin-orbit splittings in graphene but this has not yet been achieved on a semiconductor substrate. For graphene/SiC(0001), Au intercalation yields two phases with different doping. Here, we report the preparation of an almost pure p-type graphene phase after Au intercalation. We observe a 100 meV Rashba-type spin-orbit splitting at 0.9 eV binding energy. We show that this giant splitting is due to hybridization and much more limited in energy and momentum space than for Au-intercalated graphene on Ni. In a future spintronics, graphene could be used in various ways: as efficient leads for spin currents due to low spin-orbit interaction and large spin relaxation lengths [1], as ferromagnetic graphene due to a proximity magnetization [2], as spin filter when a large Rashba type spin-orbit interaction and a band gap at the Dirac point are imposed by a substrate [3], or when the spin-orbit interaction is strongly enhanced by covalently bonded adatoms [4]. The Rashba effect on a graphene Dirac cone leads to a complex band topology with a gapped and an ungapped cone similar to spinless bilayer graphene [5,6]. It has been shown by ARPES that for Au-intercalated graphene/Ni(111) the Dirac cone is intact [7] and gapless [8]. A giant Rashba splitting of ∼ 100 meV was found for graphene/Au/Ni(111) [9].On light elements, the spin-orbit splitting of graphene cannot be resolved in experiment, this is the case with Ni(111) and Co(0001) [10,11]. Where, instead, an unexpected Dirac point forms at higher binding energy as part of a pair of bonding and antibonding Dirac cones [12]. These cones have been found to be highly spin polarized by the exchange interaction from the substrate [13,14].The interfaces of graphene with metals are structurally rather well defined, and the intercalation of Au is driven by the transition metal substrate. The creation of the spinorbit interaction in the graphene is well understood based on spin-dependent hybridization with d-states of the substrate. This is the case for graphene/Au/Ni(111) [9] and is confirmed indirectly by the vanishing spin-orbit splitting of the graphene-Bi interface because atomic numbers of Bi and Au are similar but hybridization with d-states is possible only for Au [15]. A similar hybridization causes the ∼ 50 meV spin-orbit splitting for graphene/Ir (111) [16].To make the aforementioned effects useful for spintronics, they have to be confirmed on an insulating substrate such as SiC because otherwise the metallic substrate will dominate charge and spin transport. On bare SiC, graphene does not show an enhanced spin-orbit splitting as has been clarified by spin-resolved photoemission [17]. Recently, a spin-orbit splitting of 17 meV has been measured for graphene on the semiconductor WS 2 [18].In the context of a possible p-doping of graphene by metals, the intercalation of Au under graphene on SiC has already been studied [19,20], which serves as starting point for our search for enhanced spin-orbit splittings in this s...