磷化铟纳米管的合成

One-dimensional structures, such as semiconductor nanowiresor carbon nanotubes, are attractive building blocks for nanoelectronics. In contrast to organic molecules and quantum dots, theycan relatively easily be manipulated and contacted to formfunctional circuits.1 The electronic structure of semiconductingnanowires is determined by the chemical composition and thediameter and additionally these wires can be n-type or p-typedoped2 Moreover, atomically abrupt heterojunctions have beensynthesized in a nanowire,3 and strong quantization effects in aone-dimensional nanowire were established by the incorporationof a quantum dot in the wire.4 However, the confinement fornanowires with a homogeneous composition, which still havemacroscopic dimensions in one direction, is clearly less than forquantum dots. The confinement effect in one-dimensional structurescan be enhanced by the formation of tubes instead of wires. Recentlythe synthesis of transition-metal chalcogenide nanotubes has beenreported.5 They have a pseudographitic morphology, and theelectronic properties are determined by the diameter and the chiralityof the tubes, in analogy to carbon NTs.6 Therefore, it is difficult topredetermine the electronic properties of tubes of these materialclasses.

In this contribution the synthesis of crystalline and opticallyactive InP nanotubes is reported. The InP tubes have the zinc blendestructure and therefore represent a new class of tube materials. Thetubes are grown via the VLS (vapor-liquid-solid) laser ablationmethod, analogous to nanowire growth.7 The synthesis was carriedout without the use of a template. The ablation setup and synthesisconditions are similar to those reported in ref 8. The beam of anArF laser (A = 193 nm, 100 mJ/pulse, 10 Hz) is focused on apressed InP target (density 65%). A silicon sample with a nativeoxide layer covered with an equivalent of a 2-20 A Au film wasused as the substrate. The substrate was placed on an AlO; blockat the downstream end of the tube oven, and the temperature wasmeasured 1 mm below the substrate with the use of a thermocoupleThe substrate temperature was well stabilized before the ablationwas started. After synthesis the sample was studied with electronmicroscopy. For TEM a Cu grid provided with a film of amorphouscarbon was wiped over the substrate.

When the substrate temperature was in the range 430-500 Cand an undoped InP (6N) target was used, single-crystalline InPnanowires were formed, which were grown in the [111] directionas has been reported by others.8 However, when higher temperatures(>500°C) were applied, InP nanotubes were formed. Figure lashows a top view SEM (Philips XL40FEG) image of a substrateafter growth at a substrate temperature of 515 ·C, over 95% of thedeposited material consisted of wire-like structures. A closerexamination by TEM (TECNAI TF30ST), presented in Figures 1.b-e, showed that hollow tubes were formed. Judging from thecontrast from both bright-field TEM and HAADF (high-angleannular dark field) imaging it was clear that there was no materialpresent in the core of the tubes. The tubes were terminated by aparticle (Figure 1c) which contained gold; this indicates that the tubes grow from the liquid InP-Au phase via the VLS mechanism.The diameter of the tubes was uniform along their length. Thedarker regions in Figure ld correspond to the walls of the tube andare due to diffraction originating from the crystalline nature of thistube. Upon tilting the sample with respect to the electron beam.diffraction fringes can be seen to move over the entire width ofthe tubes, implying a cylindrical shape of the crystals. Theobservation that the diffraction contrast is most pronounced in thewalls again confirms the hollow nature of the tubes. The thicknessof the wall of the nanotube shown in the HRTEM image in Figurele was approximately 4 nm. From electron diffraction and X-raydiffraction measurements it was clear that the crystal latticecorresponded to the InP zinc blende lattice. The tubes did notoxidize upon exposure to ambient air for over a month. Preliminaryphotoluminescence experiments showed emission at 590 nm uponexcition at 514 nm for tubes having a wall thickness of approximately 2 nm and a diameter of 27 nm, whereas Inp wireswith a similar diameter emit at 880 nm. This increased blue-shiftdemonstrates that the quantum confinement effect is stronger inthese tubular structures than in solid nanowires.

In the presence of dopants a different morphology, ie. nanotubespartly filled with InP crystallites, was also observed. The morphol-ogy depended on the dopants added to the InP target and on thesubstrate temperature. Figure 2 shows a TEM image of a partiallyhollow tube formed by using an InP target doped with 1 mol % Seat a substrate temperature of 513 °C. From high-resolution studiesit was clear that the dark regions correspond to crystalline InPand bright- and dark-field TEM imaging showed that there was nomaterial present at the lighter regions in the core. Table 1 shows the morphology obtained when dopants (0.1 or 1 mol %) wereadded to the target at a given temperature. With static SIMS andXPS it was shown that dopant atoms were transferred from thetarget into the wires. When a target dopant concentration of 1.0mol % with respect to InP was used, an estimated concentration of1020 atoms/cm3 was present in the wires. These macroscopicchemical analyses were done on a large number of wires andtherefore do not give information about the dopant distribution inthe nanostructures. The target dopant concentration had no influenceon the resulting morphology. At the higher temperatures tubes wereformed, and at the lower temperatures, solid wires. At intermediatetemperatures the partly filled tubes were observed.