(a) Phase images for 400-nm-thick junctions with the different dopant concentrations. (b) Cosine of the phase clearly showing that the near-surface regions of the specimens are at a negative (-type) potential.
The step in phase measured across the junctions as a function of the crystalline specimen thickness. The specimens have been examined after FIB preparation at 30 and 8 kV and after annealing in vacuum. (a), (b), and (c) refer to the specimens with dopant concentrations of , , and , respectively.
(a) The electrically inactive thickness as a function of the dopant concentration in the specimens. (b) The electrically inactive thicknesses measured in the differently doped specimens as a function of the FIB operating voltage. Specimens that were annealed at low temperature have been indicated as having been prepared at 0 kV.
2D simulations of the potential in specimens with dopant concentrations of , , and . The specimens are assumed to be perfect in the simulations with a 25-nm-thick amorphous layer present on each surface. A layer of charge has been placed between the crystalline and amorphous regions and has been varied to assess the effects of the potentials within the differently doped regions.
Potential profiles extracted from across the simulations of the differently doped junctions for sheet charges of (a) 0, (b) , and (c) .
Potential profiles extracted from across the simulations for the differently doped junctions with 25-nm-thick amorphous layers containing a trapped negative charge of .
The electrically inactive thickness as a function of dopant concentration in symmetrical junctions for simulated perfect specimens and for specimens prepared using a FIB operated at both 30 and 8 kV and measured experimentally. The inactive thicknesses for the annealed specimens are also shown.
The electrically inactive thickness measured (in units of nanometers) for each of the differently doped specimens for each of the different preparation methods. The annealed specimens can be prepared using either 30 or 8 kV Ga ions to achieve the same experimental result. The experimental error in each case is ±15 nm.
The electrically inactive thicknesses (in units of nanometers) after annealing in the differently doped specimens measured experimentally and from simulations using either a sheet charge of on the specimen surface or a 25-nm-thick layer containing a positive charge (volume) of layers. The error in the experimental profiles is ±15 nm.
Article metrics loading...
Full text loading...