Low energy scanning transmission electron beam induced current for nanoscale characterization of p–n junctions

Abstract

Electron beam induced current (EBIC) at p–n junctions can be measured in high spatial resolution using a thin lamella geometry, where most incident electrons transmit the sample. We explore the case of low excitation energies in a wedge‐shaped lamella geometry to increase resolution in a controlled way. We compare a sample with high (Si) and low (manganite‐titanate heterojunction) diffusion length and use Monte Carlo based simulations as a reference. It is shown that the EBIC signal obtained from the Si junction vanishes below a thickness of 300 nm, whereas this happens at 80 nm in the PCMO–STNO junction. This allows for achieving an EBIC resolution of better than 50 nm for the latter system. The observed fundamental differences between the silicon and the perovskite junction are discussed in terms of preparation induced ‘dead’ layers and surface recombination.

Contribution of Jahn-Teller and charge transfer excitations to the photovoltaic effect of manganite/titanite heterojunctions

“…Under polychromatic illumination, the linear temperature dependence of V OC in the entire temperature range and the similar values of the slopes in PCMO/STNO and CMO/STNO junctions ( figure 4(a)) point to a dominant contribution to the photovoltaic effect, which originates from excess carriers generated in the STNO. This is in accordance with recently published low-energy scanning transmission electron beam-induced current observations for PCMO/STNO junctions [33]. In addition, a strong increase of the short-circuit current due to STNO excitations was observed in [20].…”

Microscopic electronic and structural analysis of femtosecond laser sulfur hyperdoped silicon

“…Further experiments should include EBIC experiments at very high spatial resolution in order to better resolve defective regions. A previously described approach of low‐energy scanning transmission EBIC () performed on thin samples might be appropriate, although FIB prepared silicon lamellas suffer from severe surface recombination which needs further sample preparation development.…”

“…This complication does not occur in a cross‐section geometry (Fig. ), although a quantitative analysis here is complicated by two major effects, that is, the partial transmission of electrons due to the finite size of the hillocks () as well as the effect of surface recombination, which is strong at FIB prepared surfaces . Transmission losses are not severe in these samples (except in parts very close to the surface) and surface recombination at the FIB‐prepared surface should be rather homogeneous.…”

Recombination and Charge Collection at Nickel Silicide Precipitates in Silicon Studied by Electron Beam‐Induced Current

“…A peculiarity of EBIC measurements at FIB‐prepared structures is recombination processes near the surfaces exposed to the ion beam during FIB preparation. [ 38 ] In addition to an enhanced surface recombination velocity drastically reducing the effective excess carrier lifetime, [ 30 ] the existence of a “dead layer” has been shown by Peretzki et al, [ 39 ] that is, a layer with a width of several 10 nm (depending on the material and preparation conditions) not contributing to the EBIC signal at all. This situation helps to establish a spatial resolution in EBIC experiments sufficient to resolve the charge‐collecting structures in the submicrometer range if small acceleration voltages are used, where a substantial part of high‐energy‐electron multiple scattering—and hence electron–hole pair generation—is inside the dead layer.…”