Do we need a new methodology for GaAs passivation?

Previous results on the GaAs:native oxide interface are reviewed and their implications to present attempts at GaAs passivation are discussed. Large band gap semiconductors and the use of sulfur and Sb termination as a means to passivate the surface are discussed. Although sulfur treatments do not lead to ideal (flatband) electrical termination, this treatment exhibits greatly reduced surface recombination. Sb gives an ideal termination which results in the surface Fermi level close to the flatband condition. Both of these treatments have failed as practical passivating processes because they have insufficient chemical stability against the atmosphere. A new methodology for passivating compound semiconductors is presented in which two overlayers are used. In this approach, the first overlayer defines the surface electrically and the second provides long term protection.