No abstract

The passivation of the interface between Ge and the gate dielectric is a critical issue for the integration of Ge into next generation CMOS devices. GeO2 has recently garnered a lot of interest, but there is always a trade-off between low interface state densities and a low equivalent oxide thickness. In this paper we investigate the S-passivation of the Ge gate stack in which only 1 monolayer of S is needed in order to improve the interface properties of the gate stack. S-passivation is achieved via exposure of the clean Ge(100) surface to H2S. The high-k dielectric is deposited via atomic layer deposition. We show that the oxidant precursor type (H2O versus O3) will result not only in different growth behavior but also in different interface properties. The H2O based process results in low defect densities at the valence bandedge, whereas the O3 based process results in low defect densities at the conduction bandedge.

High mobility channels are currently being explored to replace the Si channel in future technology nodes. However, until now the promising bulk properties are very difficult to translate into reality due to a bad passivation of the interface between the gate stack and the substrate. In this paper we want to emphasize that gate stack passivation is the result of a complex interplay between the surface treatment and the ALD process. During ALD on GaAs, the removal of surface oxides is observed. However, this effect does not lead to a good passivation. For Ge, the S-passivation of the interface is studied in combination with different high-κ. It is clear that the Ge/S/Al 2 O 3 interface is superior to the Ge/S/ZrO 2 interface.