Hydrogen-terminated diamond exhibits a high surface conductivity (SC) that is commonly attributed to the direct action of hydrogen-related acceptors. We give experimental evidence that hydrogen is only a necessary requirement for SC; exposure to air is also essential. We propose a mechanism in which a redox reaction in an adsorbed water layer provides the electron sink for the subsurface hole accumulation layer. The model explains the experimental findings including the fact that hydrogenated diamond is unique among all semiconductors in this respect.

The electron affinity (EA) $\chi$ of single crystal diamond (100) is determined as a function of hydrogen and oxygen coverage by a combination of work function and photoemission experiments. For the fully hydrogenated (100)-$(2\mathrm{}\times 1):\mathrm{H}$ surface an EA of $-1.3\hspace{0.5em}\mathrm{eV}$ and for the oxidized surface C(100)-$(1\mathrm{}\times 1):\mathrm{O}$ $\chi =+1.7\mathrm{}\hspace{0.5em}\mathrm{eV}$ are obtained. These are the lowest and the highest electron affinities, respectively, ever reported for any diamond surface. The variation in $\chi$ with O and H coverage is well described by a simple dipole model provided that the depolarization is properly taken into account for high adsorbate densities. This analysis favors the bridge position (etherlike) for oxygen on C(100). By mixing H and O adsorbates on a microscopic scale the EA of C(100) can be adjusted at will over 3 eV between the extreme values without jeopardizing the chemical passivation of the diamond surface afforded by H or O termination.