Publikationsrepositorium - Helmholtz-Zentrum Dresden-Rossendorf

Off-axis electron holographic tomography (EHT) has been successfully applied to reveal the 3D structure of III-V semiconductor core-shell nanowires (NWs) [1,2]. The technique probes the phase shift of an electron wave transmitted through such a NW that is proportional to the NWs projected electrostatic potential. Thus, a tilt series of phase images (projected potentials) can be used as input to compute a 3D tomogram of the electrostatic potential by tomographic reconstruction algorithms. Typically, the recovered 3D potential is dominated by the mean inner potential (MIP), which is related to the materials composition. Consequently, space charge potentials determining for example the electric properties, e.g., at interfaces or pn-junctions in semiconductors [2] may be superimposed by MIP variations caused by compositional changes within the heterostructures.
Here, we show on the example of a GaAs/AlGaAs core-multishell NW, how the space charge potentials can be uncovered from materials contrast (MIP) by determining the latter independently: To this end, high-angle annular dark-field (HAADF) STEM tomography was applied in addition to EHT on the same NW. STEM tomograms provide solely materials contrast that depends exponentially on the atomic number. Fig. 1 compares both methods in terms of the relation between reconstructed signal and projected property, exemplary for three different tomogram regions identified as pure Au, GaAs and AlGaAs: In case of EHT between the reconstructed potential and the MIP, and in case of STEM tomography between the reconstructed intensity and the atomic number. The latter relation enables converting the STEM tomogram in units of (mean) atomic numbers.
Tilt series were acquired from -70° to +71° with 3° tilt steps in holography mode, and from -68° to +68° with 2° tilt steps in STEM mode. Since phase images of axially scattered electrons are used for EHT, it suffers much more from diffraction contrast than STEM tomography (high-angle scattering). Consequently, only 39 projections could be used for tomographic reconstruction in the case of EHT compared to 68 in the case of STEM tomography. For this reason, resolution and contrast in the 3D potential are slightly lower than in the STEM tomogram, which can be seen on the cross-section of the NW in Fig. 2. Nevertheless, the core-shell structure, the ca. (5-10) nm thick GaAs shell acting as quantum well tube (QWT), and unintended Al segregations are clearly resolved in both cases. Last but not least, longitudinal slices (Fig.3) exhibit clear differences of both tomograms that strongly suggest additional local space-charge related potentials to be investigated in greater detail in a next step.
References:
[1] A Lubk, D Wolf, P Prete, N Lovergine, T Niermann, S Sturm and H Lichte, Phys. Rev. B 90 (2014) p. 125404.
[2] D Wolf, A Lubk, P Prete, N Lovergine and H Lichte, J. Phys. D: Appl. Phys. 49 (2016) p. 364004
[3] We thank N Lovergine of University of Salento, Lecce for provision of the samples.
[4] We thank the group of Michael Lehmann at TU Berlin for access to the TEM FEI Titan 80-300 Berlin Holography Special.
[5] DW acknowledges financial support within the European Union"s Horizon 2020 research and innovation program under Grant Agreement No 688072 (Project IONS4SET). AL has received funding from the European Research Council (ERC) under the European Union"s Horizon 2020 research and innovation programme (grant agreement No 715620).