Conductive-probe atomic force microscopy and Kelvin-probe force microscopy characterization of OH-terminated diamond (111) surfaces with step-terrace structures

Abstract

We characterized OH-terminated diamond (111) surfaces which show step-terrace (ST) and bunching-step (BS) regions from contact potential difference (CPD) and current to investigate the relationships between surface topography and generation of interface states on the surfaces. The OH-termination was performed using water vapor annealing. The CPD and current measurements were conducted by conductive-probe AFM and Kelvin-probe force microscopy. The CPD and current were highly correlated with the surface topography. The I–V characteristics shows typical ideality factors (n) of about 1.5 and 2.0 in the ST region and the BS region, respectively. As the n were higher than that of an ideal Schottky contact (n = 1.0), we concluded that a metal insulator semiconductor diode structure, whose n increases as its interface state density increases, was formed. Considering that step density in the BS region was much higher than that in the ST region, the steps seemed to have generated the interface states.

Grain and Grain Boundary Conduction Channels in Copper Iodide Thin Films

“…When a defect is saturated by an adatom (like oxygen), the local properties change. [ 30,37 ] It could be any kind of element, but our measurements after oxygen plasma treatment show similar loss of surface conductivity as for storing the samples in atmosphere for a longer period of time while argon treatment does not seem to diminish the surface conduction, which suggests oxygen being the adatom which changes the electrical configuration. This degradation of the grain surfaces occurs more rapidly in comparison to the GBs because the defect density at the grains is lower and a larger surface can be attacked by oxygen, but also because the 2DEG on the GBs attracts water molecules with a permanent dipole moment because of electrostatic interaction, [ 27,33,36 ] which could slow down the diffusion of adatoms to the defects.…”

“…The mask derived from the topography data does implicate where the GBs are located. The measured current is in the range of ±8 μA which is rather high for cp‐AFM measurements [ 24,27,30–32 ] but within the expected range for copper iodide. [ 16 ] The current is localized, as one can clearly separate light blue and dark red regions from gray regions which do not exhibit measurable current at all.…”

“…Commonly, a line correlation of the current and topography data would be performed [ 30 ] (Figure S6, Supporting Information), which would neglect most of the acquired data points. By using our methodology, all data points from the 3D data are evaluated and correlated for the entire measured voltage range.…”

“…This is due to the very high conductance of this material resulting in a significantly smaller resistance in comparison to the investigated sample-probe interface across which the main potential drop occurs. Commonly, a line correlation of the current and topography data would be performed [30] (Figure S6, Supporting Information), which would neglect most of the acquired data points. By using our methodology, all data points from the 3D data are evaluated and correlated for the entire measured voltage range.…”

“…Here, the defect density is not as large because the surface energy at (111) is very low and we expect a rather closed layer [29] with only one dangling bond per copper atom pointing out of plane and interface defect band might be formed but not an 2DEG inversion layer. [17,30] In addition, a (À111) surface has a significant work function difference [18] and therefore a different bridging conductivity can be assumed. Due to the comparably (in relation to the GBs) low defect density, hole depletion also occurs here, but not inversion and formation of a 2DEG (otherwise a current flow in both voltage direction could be measured).…”