Ludwig Feigl1 Philipp Schroth1 2 Julian Jakob1 Seyed Mohammad Mostafavi Kashani2 Jonas Vogel2 Arman Davtyan2 Ulrich Pietsch2 Tilo Baumbach1

1, Karlsruhe Institute of Technology, Eggenstein, , Germany
2, University of Siegen, Siegen, , Germany

The small footprint of III-V nanowires allows for their epitaxial defect-free integration on Si substrates. Their large aspect ratio makes them suitable for three-dimensional chip design[1], photovoltaics[2] and sensorics[3]. In order to create any device structures, heterostructures have to be fabricated, e.g. by the deposition of a shell layer[1-3].

We are growing (In,Ga)As core-shell nanowires in a portable MBE chamber based at the Karlsruhe Institute of Technology that is specially designed for time-resolved in-situ X-ray experiments during growth, under true MBE conditions[4]. Using X-ray diffraction (XRD), we are sensitive to the nanowire shape, phase composition, strain and relaxation processes[5]. In order to resolve the temporal evolution of these characteristics during nanowire and shell growth, we utilized the high-brilliance synchrotron radiation at beamline P09 of PETRA III, Hamburg.
During in-situ XRD experiments the crystal structure is monitored by collecting partial reciprocal space maps (RSMs) around the asymmetric (311) and (220) zincblende and (10.3) wurtzite Bragg reflections of In0.30Ga0.70As, GaAs and Si with a temporal resolution of about three minutes. Additionally, high resolution three-dimensional in-situ RSMs are done both at growth temperature and at room temperature before and after shell growth. The X-ray beam was tuned to a spot size of several micrometers which allows analyzing an ensemble of nanowires to address the statistical character of the growth process.

During the fabrication of the shell, we observed the gradual appearance of additional (In,Ga)As Bragg peaks. At the same time, the Bragg intensities measuring the GaAs nanowire core were affected. In particular, a decrease of the peak amplitude is attributed to local distortions caused by the introduction of misfit strain. Additionally, the decreasing crystal truncation rod intensity can be interpreted by a reduction of the stacking fault density and the changing ratio of the phase selective Bragg peaks indicate a decreasing wurtzite phase fraction. This demonstrates the sensitivity of the crystal structure of GaAs nanowires when used as core for shell growth. Furthermore, our time-resolved in-situ investigations revealed that this phase change occurs already during the first stage of shell growth. Such insights into the temporal evolution of the microstructure during shell formation are beneficial for future attempts to tailor the crystal structure of core-shell nanowire heterostructures.

[1] Tomioka et al. IEEE J. Sel. Top. Quantum Electron. 17, 1112 (2011)
[2] Dimakis et al. Nano Lett. 14, 2604 (2014)
[3] Cui et al. Science 293, 1289 (2001)
[4] Slobodskyy et al. Rev. Sci. Instrum. 83, 105112 (2012)
[5] Schroth et al. Phys. Rev. Lett. 114, 055504 (2015)