Fig 5.3a shows a CdTe NW array, generated using Tsub = 520°C, Tsource = 550°C, P = 25 Torr and t = 30 mins. An example of an individual NW is shown in Fig 5.3b – it has a
diameter of ~ 250 nm that is constant along its axis (see Section 5.3.5.2 for a discussion as to why this diameter is larger than the initial nanodot diameter), a length, l, of ~ 10 μm
(determined from a lower magnification image) and a spherical droplet at the tip (confirmed below to be the Au catalyst). Its growth axis is not perpendicular to the substrate.
Fig 5.3: a) Secondary electron SEM image of a CdTe NW array grown by subliming CdTe at 550°C for 30 mins under 25 Torr nitrogen onto a Au-nanodot covered Mo substrate held at 520°C. b) Image of an individual NW with a spherical droplet present at the tip.
77 A distribution of NW lengths and diameters was obtained, with some exceeding 20 μm in length and having diameters in the range 50 - 500 nm. The distribution of NW diameters is considered to be governed by the distribution of the Au catalyst drop diameters. Fig5.4 shows the relationship between NW length and diameter for approximately 70 NWs; typically, the thinner NWs grew to greater lengths than the thicker NWs. The orientation of NWs with respect to the substrate appears to be random (see Section 5.4). The NW density (in the range 106 – 107 cm-3, estimated from an average of measurements taken at different regions of the substrate) is roughly two orders of magnitude lower than the initial density of Au dots
(~ 109 cm-3, inferred from results from Section 5.2 for annealing Au films at 360°C). A rough semi-continuous thin film (~ 500 nm thick, confirmed to be CdTe by EDX) was also
observed, at the base of the NWs and this is described in more detail in Section 5.3.3.
Fig 5.4: The relationship between nanowire length and diameter, measured directly from SEM micrographs for ~ 70 nanowires. Typically, the thinner NWs grew to greater lengths than the thicker NWs.
SEM/EDX microanalysis was used to determine the composition of both the stem and tip of a typical NW. A spectrum taken from the stem of a NW (Fig 5.5a) clearly shows Cd and Te Lα1 peaks but no Au, whereas a spectrum taken from the droplet (Fig 5.5b) is dominated by a Au peak. Fig 5.5c shows EDX mapping of the Au Lα1 peak overlaid with the accompanying secondary electron image, and Fig 5.5d is a TEM image demonstrating the highly abrupt interface between catalyst and the NW (in-depth TEM analysis is provided in Chapter 6). While both the EDX spectrum and mapping indicate that the droplet is at least Au-rich, it is not clear from these experiments whether it is pure Au or else contains Cd and Te (peaks for
78 all three elements are present in Fig 5.5b). It remains possible that the apparent presence of both Cd and Te in the droplet is due to a measurement artefact such as beam spreading or the excitation of a signal from a nearby wire.
Fig 5.5: Energy dispersive X-ray diffraction (EDX) taken from a) NW stem and b) droplet at the NW tip. c) EDX mapping of the Au Lα1 peak overlaid on a secondary electron image of the NW tip. d) TEM image demonstrating the highly abrupt interface between droplet and NW.
For a more reliable chemical analysis, NWs were mechanically removed onto a Si substrate, and a number of EDX spectra were taken to investigate the spatial distribution of the elements. Fig 5.6a shows a representative example of such a NW and the analysis points. Fig 5.6c shows the corresponding atomic % of Au, Cd and Te at these points, with scan 1 being taken from the droplet, and scan 9 being taken from the furthest-most point from the droplet. Likewise, the data shown in Fig 5.6d corresponds to the higher magnification image shown in Fig 5.6b. The uncertainty assigned to the composition values was √ where is the number of counts. The following comments refer to the higher resolution data in Fig 5.6d.
At the centre of the droplet, the composition is 100.0 % Au, although it must be noted that the sensitivity of EDX is limited to 100 ppm. Towards the edge of the droplet however, both
79 Cd and Te are detected (scans 2 and 3, Fig 5.6d). Moreover, Au was detected as a major impurity in the upper 50 nm of the CdTe NW stem itself (scans 5 and 6, Fig 5.6d). These observations may be real or may be the result of artefacts, either due to: a) the orientation of the NW on the Si – in this plane of view, the Au droplet is obscuring part of the NW itself; or b) beam spreading.
Fig 5.6: EDX was taken at various points along a NW to determine the spatial variation of the atomic % of Au, Cd and Te. This was first done at low-magnification in the SEM (a), with the corresponding composition plot shown in c. Similarly, the plot in d) corresponds to the higher magnification image in b).
In order to evaluate the influence of beam spreading, the electron trajectories in a CdTe NW (diameter ~ 100 nm) lying on a Si substrate upon irradiation with an 8 kV beam, were modelled using a Monte Carlo method. From Fig 5.7 it is clear that a large proportion of X- rays were emitted from a generation volume of radius as large as ~ 75 nm. Ultimately, while beam spreading and the angle of the interface are evidently significant on the scale of the analysis intervals in Fig 5.6b, scan 1 indicates that at least the tip of the droplets comprises Au alone.
80 Fig 5.7: Monte Carlo simulation of the electron trajectories in a CdTe NW lying on a Si substrate irradiated by an 8kV electron beam.