(395f) Influence of Negative Adsorption on Hydrogen Terminated P-Type Diamond Surface

It is known that Hydrogen terminated surface of as grown intrinsic diamond has p-type useful high conductivity that closely related to surface adsorbates. The sheet resistance and the sheet carrier density, the carrier mobility of usual single crystalline hydrogen-terminated diamond (100) surface is about 10k ohm/sq and 10E13/cm2, 150cm2/Vs respectively. Especially, through the use of very simple fabricate process (do nothing other than replacing hydrogen atom for surface atom of undoped diamond) and ability to separate nano-scale conductive and insulating region used by insulating layer of oxygen-terminated surface of diamond, it is going to be important quality for nanostructured device of diamond. In fact, some microstructured devices with this conductive layer have already been made and its utility has been sufficiently proven [1,2]. So it is very important to have a sufficient understanding of a formation mechanism of surface conductive layer. Some models have been proposed to explain the origin of this surface conductivity (SC). Above all models, transfer doping model that Maier et al. proposed [3] attracts a lot of attention. In this model, electrons near the valance band maximum of diamond are transferred to an electrochemical potential (H3O+/H2) of adsorbed water layer. According to the model, the surface conductivity increases as pH decreases. But we have so far observed that the sheet carrier density of SC doesn't depend on the concentration of hydronium ions (pH) when the surface is terminated by hydrogen [4], but the concentration of negative halogen ions in the solutions, from the characterization of electrolyte-solution gated diamond FETs (SGFETs)[5]. This pH insensitivity contradicts to the transfer doping model which is based on electrochemical potential. Sheet carrier density of the SC is constantly 10E13[/cm2] order in air. Hydrogen-terminated diamond surface has surface dipoles by the electronegativity difference between C (2.1 Pauling units) and H (2.5 Pauling units) at (100) surface and the density of surface positive charge by hydrogen atoms is 6x10E13 [/cm2]. So it is reasonable that negative charged adsorbates induce minority hole accumulation by band-bending like inversion layer of MOSFET and make SC on positive hydrogen-terminated diamond surface. In this work, we have introduced large amount oxygen negative ion (O2-: to be found in regular air) by low energy electron collisional attachment on the hydrogen-terminated surface which have been changed to be highly resistive by heat treatment in vacuum, and measured the change in electrical conductivity. Figure shows an attempt to, with oxygen negative ion, bring SC to Hydrogen-terminated diamond (100) surface whose SC was lost by the heat treatment at 550K. The SC before anneal was 4.2x10E-6[S]. After this SC was lost by the heat treatment-in high vacuum, pure O2 was induced in chamber till 76 Torr at t=30 [min]. At this time, SC was just a little made a recovery, but keep low conductivity. And at t=35 min, negative oxygen ions were produced in chamber by low energy electron collisional attachment. At t=50 min, SC was made a recovery up to 3.5x10E-6[S] and keep SC after exposed to air. This result with negative charged adsorption indicates that there is phenomenon unexplained by transfer doping mechanism and possibility that negative charged molecules in air are closely linked to the origin of SC. We are going to show other evidences which support the negative charge model in our presentation. [1] H. Matsudaira et al., IEEE Elect. Dev. Lett., 25 (7), 480-482 (2004)., [2] M. Tachiki et al., Phys. Stat. Sol., 199(1), 39, (2003)., [3] F. Maier, et al., Phys. Rev. Lett. 85 (2000) 3472., [4] H. Kawarada, et al., Phys. Status Solidi A 185 (2001) 79., [5] K.S.Song et.al.,Biosens Bioelectron. (2003)19,137-140.