Beam damage of single semiconductor nanowires during X-ray nanobeam diffraction experiments

Abstract

Nanoprobe X-ray diffraction (nXRD) using focused synchrotron radiation is a powerful technique to study the structural properties of individual semiconductor nanowires. However, when performing the experiment under ambient conditions, the required high X-ray dose and prolonged exposure times can lead to radiation damage. To unveil the origin of radiation damage, a comparison is made of nXRD experiments carried out on individual semiconductor nanowires in their as-grown geometry both under ambient conditions and under He atmosphere at the microfocus station of the P08 beamline at the third-generation source PETRA III. Using an incident X-ray beam energy of 9 keV and photon flux of 10¹⁰ s^-1, the axial lattice parameter and tilt of individual GaAs/In_0.2Ga_0.8As/GaAs core-shell nanowires were monitored by continuously recording reciprocal-space maps of the 111 Bragg reflection at a fixed spatial position over several hours. In addition, the emission properties of the (In,Ga)As quantum well, the atomic composition of the exposed nanowires and the nanowire morphology were studied by cathodoluminescence spectroscopy, energy-dispersive X-ray spectroscopy and scanning electron microscopy, respectively, both prior to and after nXRD exposure. Nanowires exposed under ambient conditions show severe optical and morphological damage, which was reduced for nanowires exposed under He atmosphere. The observed damage can be largely attributed to an oxidation process from X-ray-induced ozone reactions in air. Due to the lower heat-transfer coefficient compared with GaAs, this oxide shell limits the heat transfer through the nanowire side facets, which is considered as the main channel of heat dissipation for nanowires in the as-grown geometry.

Keywords: air and He atmosphere; individual core–shell nanowires; morphological and optical changes; nanoprobe X-ray exposure; ozone-induced oxidation.

Figure 1

(a), (b) SEM micrographs of NW1 before and after exposure for 1 h, respectively. (c), (d) Normalized low-temperature hyperspectral CL line scans along the NW growth axis before and after exposure (color-coded logarithmic intensity scale), respectively, for NW1. Note that the emission intensity depends sensitively on the positioning of the NW with respect to the focal point of the parabolic mirror so that the absolute intensities of the two line scans cannot be compared directly. (e), (f) SEM micrographs of NW2 before and after exposure for 4 h, respectively. Red and pink dashed circles indicate NW sections illuminated by the peak (P) and tail (T) of the Gaussian beam. Colored arrows indicate the positions at which the EDX measurements shown in Fig. 2 ▸ were carried out. The scale bar in (a) corresponds to 1 µm and applies to all SEM images.

Figure 2

EDX spectra taken at the top, middle and bottom sections of (a) NW2 (air) and (b) NW4 (He). The color of each line profile is correlated with the color of the arrows in Fig. 1 ▸(f) for NW2 and in Fig. 3 ▸(f) for NW4. The horizontal pink lines correspond to the peak height for each element in a non-exposed NW.

Figure 3

(a), (b) SEM micrographs of NW3 before and after exposure for 1 h, respectively. (c) Normalized low-temperature hyperspectral CL line scan along the NW growth axis after exposure (color-coded on a logarithmic intensity scale) for NW3. (d), (e) SEM micrographs of NW4 before and after exposure for 3 h, respectively. Colored arrows indicate the positions at which EDX measurements were carried out. (f) Normalized low-temperature hyperspectral CL line scan after exposure for NW4. The scale bar in (a) corresponds to 1 µm and applies to all SEM images.

Figure 4

(a)–(e) 2D projections of the 111 Bragg reflection in the (, ) reciprocal-space plane for NW2 exposed in air. The two dashed curves in represent the ZB (top) and WZ (bottom) Debye–Scherrer rings. The red and pink dotted circles named P and T are explained in Fig. 1 ▸(e) and represent sub-Bragg peaks that originate from NW sections illuminated by the peak and tail of the Gaussian beam. The peaks named ZB1–ZB3 are referred to in Fig. 6 ▸(d). (f)–(j) 2D projections of the 111 Bragg reflection in the (, ) reciprocal-space plane. The red and black circles represent the positions of the Si CTR and the initial NW tilt, respectively, in and . The time at which each RSM acquisition was started is mentioned at the top.

Figure 5

(a)–(d) 2D projections of the 111 Bragg reflection of NW4, exposed under He atmosphere, in the (, ) reciprocal-space plane. The two dashed curves in represent the ZB (top) and WZ (bottom) Debye–Scherrer rings. (e)–(h) 2D projections of the 111 Bragg reflection in the (, ) reciprocal-space plane. The red and black circles represent the positions of the Si CTR and the initial NW tilt, respectively. The time at which each RSM acquisition was started is mentioned at the top.

Figure 6

(a), (b) Tilt of NW2 (air) and NW4 (He) along and , respectively. The blue hexagon represents the NW cross section and the black arrow represents the tilt direction. Red data points correspond to section P, whereas the pink ones correspond to section T of NW2. The black points correspond to NW4, exposed for 3 h under He atmosphere. (c) Intensity decay of the integrated 111 Bragg reflection for the two NWs. The red curve is an exponential fit. Error bars are within the size of the dots. (d) Variation in the c lattice parameter. The orange points correspond to the WZ reflection. The three ZB sub-peaks, colored in pink, were named ZB1–ZB3 as indicated in Fig. 4 ▸(d).