Rectifying Behavior of Electrically Aligned ZnO Nanorods

We report on the electrical alignment of ZnO nanorods and their electrical properties. The ZnO nanorods were wet-chemically synthesized, and their length and diameter were adjusted to about 200−300 nm and 15−30 nm, respectively. The nanorods were deposited onto electrode structures and directed into 200to 800-nm-wide electrode gaps by using alternating electric fields at frequencies between 1 and 10 kHz and field strengths between 106 and 107 V/m. The nanorods align parallel to the electric field lines and make electrical contact with the gold electrodes. Clear photoresponse to 366-nm ultraviolet light irradiation was demonstrated. The current−voltage characteristics of the aligned rods are strongly nonlinear and asymmetrical, showing rectifying, diode-like behavior and asymmetry factors up to 25 at 3-V bias voltage. The continuous downscaling of the feature sizes of microelectronic devices well below 100 nm1 has initiated an enormous amount of research activity on alternatives to traditional electronic components such as complimentary metal oxide semiconductor (CMOS)-based field-effect transistors and lithographically defined interconnects. The demand for alternative device concepts is related to the expectation that downscaling of conventional CMOS devices will reach physical limits within the next few years.2 For the patterning technology, optical lithography will still be the method of choice for mass production because of its high throughput capabilities.3 However, the development of optical lithography tools for continuously decreasing feature sizes will become increasingly expensive. Thus, radical new device and fabrication concepts that offer high performance and low price are required. A current trend is to invent device concepts based on unique nanostructures of materials such as insulators, semiconductors, metals, or combinations of these materials. Nanorods and tube-shaped nanostructures have been fabricated recently using cheap and simple wet-chemical and physical synthesis methods. Semiconducting tubes and nanorods have been made from CdSe, CdTe, ZnO, and others, and they have shown a large variety of properties such as chemical sensing,4 luminescence,5 field effects,6 lasing,7 and photoresponse.8 Synthesis concepts are well established; however, there is less experience to date on the alignment of nanostructures to build working electronic devices or to integrate them into complex architectures.9 Unconventional lithography approaches such as microcontract printing10 or dip-pen lithography11 have the potential to control nanostructure assembly. The funtionalization of carbon nanotubes toward the directed self-assembly and alignment of more complex nanostructures has already been demonstrated.12 Here, we describe the successful alignment of wetchemically synthesized, semiconducting ZnO nanorods in the suspension phase using an electric field to place them onto predefined electrodes. We also demonstrate their unique properties such as photoresponse and current rectification. Synthesis. ZnO nanorods were prepared according to a method described earlier.13 Briefly, 29.5 g (0.13 mol) of zinc acetate dihydrate was dissolved in 125 mL of methanol at 60 °C. Then, a solution of 14.8 g (0.23 mol) of potassium hydroxide in 65 mL of methanol was added. The reaction mixture was stirred for several days at 60 °C. During this time, the nanoparticles precipitate. The length and the width of the resulting nanorods depend on the reaction time. The length of the nanorods increased considerably with longer reaction times, but the width of the nanorods grew only slightly. The precipitate was washed with methanol and centrifuged (5500 rpm, 30 min), and the resulting gel was redispersed in chloroform. The nanoparticles were diluted with a mixture of ethyleneglycol/water (2:1). Figure 1 shows a TEM image of ZnO nanorods after a reaction time of 3 days. The nanorods have a diameter of 15 to 30 nm and a length of 200 to 300 nm. * Corresponding author. E-mail: [email protected]. Phone: +49 711 5858 212. Fax: +49 711 5858 484. † Sony Corporate Laboratories Europe. ‡ University of Hamburg. NANO LETTERS 2003 Vol. 3, No. 8 1097-1101 10.1021/nl034240z CCC: $25.00 © 2003 American Chemical Society Published on Web 06/24/2003 Electrical Alignment. The manipulation and alignment based on dielectrophoretic forces in an alternating electric field were demonstrated for a number of different structures such as DNA molecules,14 gold nanowires,15 and carbon nanotubes16 as well as for metal and semiconducting nanoparticles.17 In all cases, the electric field induces charge separation, and the resulting polarization generates a dipole moment, which aligns the structure to an energetically favorable orientation (i.e., parallel to the field lines). In addition, if the field distribution is nonuniform, then a dielectric force that points into the direction of the highest field density acts on the polarized structure. The dielectrophoretic force, F, is given by the following expression: where R is the effective polarizability of the structure, E is the electric field strength, ∇ is the gradient vector operator and, ν is the structure volume. This force is responsible for the movement of the polarized structure toward the electrode edges where the electric field density has the highest value.18 The alignment force is frequency-dependent because the polarizability of the nanostructure and the surrounding solvent changes with frequency.19 We have chosen the frequency range between 1 and 10 kHz where the polarization of the nanorods dominates the polarization of the solvent molecules so that effective electrical alignment of the ZnO nanorods is possible. At much higher frequencies, the aligning force vanishes again. An HP source measure unit20 was used to generate the electric field at various strengths and frequencies. The sample was mounted on a chip carrier and wire bonded using 25-μm-thick aluminum wires to provide electrical connections. The electrode structures were fabricated on 400-nmthick silicon oxide using standard electron beam lithography and a lift-off technique. The electrode material was thermally evaporated chrome (5 nm) and gold (50 nm) on top. Typically, after applying one drop of approximately 1 mM concentrated ZnO nanorods in ethylene glycol/water to the electrode structure, the field was switched on. Depending on the gap size, the driving voltage was gradually increased from zero to about 6 Vpp to 20 Vpp (peak-to-peak) so that the electric field strength was in the range between 106 and 107 V/m at 1 to 10 kHz. The field was applied for about 20 s; subsequently, the sample was washed using one drop of ethylene glycol. After this, the surface was dried using compressed air, and finally, the field was switched off. Figure 2 shows a scanning electron microscopy (SEM) picture of a sample that was treated using the described method. The gap size between the interdigitated finger electrodes was about 800 nm, and the applied voltage was 20 Vpp at 1 kHz, corresponding to a field strength of 1.25 × 107 V/m. The nanorods obviously follow the electric field direction by aligning parallel to the flux lines. Nanorods on top of the electrodes are not aligned because the electric field strength and therefore the resulting force is zero at these spots (eq 1). To contact both ends of the nanorods to electrodes, the alignment experiment was repeated on electrodes with smaller gaps of about 200 nm. Figure 3A shows an SEM picture of an electrode array (sample 1) after one drop of the ZnO nanorod solution at a concentration of approximately 1 mM was applied and the voltage was increased to 5 Vpp between electrodes 1 and 2, corresponding to a maximum field strength of 1.25 × 107 V/m at 10 kHz. The operating frequency was increased to 10 kHz because some of the 200nm-gap electrodes tend to get damaged at 1 kHz. From the SEM picture, it is obvious that most of the nanorods were collected in the gap between electrodes 1 and 2; a few rods Figure 1. TEM image of ZnO nanorods after a reaction time of 3 days. F ) 1 2 Rν∇|E| (1) Figure 2. Electrically aligned ZnO nanorods between 800-nm spaced, interdigitated gold electrodes on silicon oxide. The electrode fingers on the bottom and on the top were on the same potential. The applied electric field strength was 1.25 × 107 V/m, and the frequency was 1 kHz. 1098 Nano Lett., Vol. 3, No. 8, 2003 were located in other gaps. However, the alignment is quite disordered because a large fraction of the rods are not aligned parallel to the field direction. In another experiment, the concentration of the ZnO nanorod suspension was reduced by a factor of 2 while keeping the electrical alignment parameters fixed. The SEM picture in Figure 3B (sample 2) shows fewer nanorods assembled in the gap than for the previous case. In a third (sample 3), fourth (sample 4), and fifth experiment (sample 5), the initial ZnO nanorod concentration was diluted by a factor of 3. These samples show only very few nanorods assembled between the connected electrodes 1 and 2 (Figure 3C/sample 3 and Figure 3D/sample 5). These results clearly demonstrate that the ZnO nanorods can be easily aligned in electric fields, which oscillate at frequencies between 1 and 10 kHz. The number of rods can be controlled by the concentration of the ZnO nanorod suspension. Electrical Characterization. The aligned ZnO rods were electrically characterized by using an HP source measuring unit21 in a two-probe configuration and in voltage bias mode. The resistance of the nanorod assemblies, Rdark, was high, on the order of about 1-4 GΩ at a bias of 1 V. (The baseline was about 10 GΩ.) As shown in Table 1, there is just a weak scaling of Rdark with the number of nanorods. The high resistivity is governed by the wide ZnO band gap of about 3.3 eV, corresponding to a wavelength of about 373 nm.22 The conductivity can be enhanced by doping, temperature increase, or ultraviolet (UV) light-induced photoconductivity. Table 1 summarizes the resistance, RUV, of the investigated nanorod assemblies under 366-nm UV light irradiation at about 0.1 W/cm2. The resistance of all samples decreased significantly as the UV light was switched on at a bias voltage of 1 V. The estimated resistance per rod under UV light was about 1 GΩ at a bias of 1 V, which includes also the contact resistance. The change in resistance ∆R(U) ) Rdark(U) RUV(U) increased with increasing bias voltage. Figure 4 shows the time-dependent photoresponse for one to two contacted nanorods (sample 5, Figure 3D). A control Figure 3. Electrically aligned ZnO nanorods between 200-nm spaced gold electrodes. The applied electric field strength was 1.25 × 107 V/m, and the frequency was 10 kHz. The following concentrations were used: (A) ZnO nanorod stock concentration of approximately 1 mM (sample 1); (B) 50% diluted stock in ethylene glycol/water (sample 2); (C) and (D) 66% diluted stock (samples 3 and 5). Table 1. Electrical Data of ZnO Nanorod Assembly Samples