Characterization and ohmic contact of hydrothermally synthesized vertical ZnO and Ag/ZnO nanowires

Vertically aligned ZnO nanowire arrays were synthesized by two-step hydrothermal method. ZnO seed layers were prepared on substrate by using anhydrous ethanol and zinc acetate dihydrate solution, followed by the generation of ZnO nanowire arrays by low-temperature liquid-phase hydrothermal methods. The ZnO nanowire arrays were prepared under different conditions to compare the effects of growth conditions on the morphology of ZnO nanowires, in order to explore the optimal growth conditions for ZnO nanowire arrays used in semiconductor device. The morphological changes of ZnO nanowire arrays grown under different conditions were systematically analyzed by SEM, XRD and other characterization means. The results show that the seed solution concentration, growth solution concentration, doping concentration and growth time all have certain effects on the morphology of ZnO nanowire arrays. Besides, the Ag/ZnO ohmic contact were investigated, the optimal annealing temperatures of 450 °C was obtained.


Introduction
Zinc oxide (ZnO) is an important direct bandgap n-type semiconductor material with a wide bandgap of 3.37 eV at room temperature, similar to another important semiconductor material GaN (3.39 eV). However, the exciton binding energy of ZnO reaches 60 meV, which is much higher than GaN (25 meV) [1]. Besides, ZnO materials have good stability in ambient environments. As a consequence, ZnO is being anticipated as the next generation functional nano-material towards remarkably diversified electronic and optoelectronic applications [2].
The preparation routes of ZnO mainly include the following three types of methods: vapor-phase method (sputtering, vacuum evaporation, ion-body CVD and vapor-phase thermal decomposition, etc), solid-phase method (mechanical grinding, crushing method and solid-phase reaction method, etc) and liquid-phase method (deposition, sol-gel method, template method and hydrothermal method, etc.) [3]. Among the above methods, the vapor-phase deposition method can change the morphology and size of ZnO structure by changing the heating temperature and reaction time, and synthesizes the advantages of good dispersion of crystals, narrow particle size distribution, consistent orientation and good denseness. However, the liquid phase method has some essential defects, such as the high cost of processing equipment, the harsh requirements for the growth environment, and serious environment pollution [4]. Solid-state synthesis of ZnO materials has become one of the most widely used method because of its low cost. However, the powder prepared by this method is not fine enough, and the purity of the required raw materials is high. In addition, impurities are easily mixed in the growth process, so that complex treatment is required in the subsequent process [5,6]. Compared with the other two methods, the liquid phase method has the advantages of high purity, good dispersion and low cost. More Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
importantly, the synthesized material has high surface activity and is easy to realize crystal doping. Inhence, the liquid phase method is the most widely used method for preparing zinc oxide materials in the laboratory at present [7,8]. Among the various methods, preparing the seed layer by dropcasting and the fabrication of ZnO nanowires by hydrothermal growth is a simpler and inexpensive method [9]. There are many pevious reports about the influence of single factor on the growth of ZnO nanowires. However, many constraints affect the quality and morphology of the ZnO nanowires such as precursor material and concentrations, growth solution concentration, temperature, and duration of growth process. Herefore, it is necessary to systematically study the influence of these factors to optimize the growth conditions of ZnO [10].
One of the simple routes of preparing ZnO seed layers is the reaction of zinc acetate dihydrate (ZAD) with ethanol solution [11]. In this paper, the seed layer was prepared by those two methods followed by ZnO nanowire array grown by hydrothermal method. Besides, the effects of seed solution concentration, growth solution concentration, growth time and aluminum doping on the morphology of ZnO nanowire arrays were compared, and the optimal conditions for the preparation of ZnO nanowire arrays suitable for gas-sensitive sensor devices by hydrothermal method were finally determined. Additionally, metal/ZnO contacts play important roles in device performance. For ohmic contact, it was reported that a non-alloyed ohmic contact with low specific contact resistance is preferred for shallow junction and low-voltage devices. In this paper, we also studied the conditions for the formation of a good ohmic contact between high conductivity silver metal with ZnO nanowires.

Experimental material growth procedures
ZnO nanowire arrays were generated on silicon substrates using a low-temperature liquid-phase hydrothermal method. The raw materials used in this experiment include: zinc acetate dihydrate (Zn(CH 3 COO 2 )·2HO 2 ), ethanolamine (HN 2 (CH 2 ) 2 OH), anhydrous ethanol (C 2 H 5 OH), zinc nitrate (Zn(NO 3 ) 2 ), hexamethyltetramine (HMTA) [12]. After the solution was naturally cooled to room temperature, the sample was removed, thoroughly washed several times with deionized water, and dried in air at 60°C to finally obtain the ZnO nanowire arrays.

Preparation of seed layer
Absolute ethyl alcohol was used as the solvent, ethanolamine was used as the stabilizer, and the zinc acetate dihydrate and the ethanolamine are dissolved in equal molar ratio. After the seed solution was prepared, it was stirred vigorously for 1 h in a water bath at 60°C by a magnetic stirrer to form a uniform and stable solution [13]. In the following experiments, we prepared three concentrations of seed solution, which were 0.025 mol l −1 , 0.05 mol l −1 and 0.075 mol l −1 . We selected the 0.05 mol l −1 seed solution to prepare the seed layer on the silicon substrate by dip-coating method and spin-coating method, to compared the quantity of seed layer formed by the two methods. In the dipcoating step, a clean silicon wafer is used as the substrate, and the silicon wafer is completely immersed in the seed solution and then vertically pulled out, and then blown dry with a stream of argon. In the spin-coating method, a dropper is used to put the seed droplet pattern on the substrate, and the rotation speed of the spin coating machine is set at 2000 r min −1 for 30 s. The two substrates, now covered with a film of zinc acetate crystallites, after heated to 300°C in air for 25 min [14]. Figure 1 shows the Atomic Force Microscope (AFM) image of seed layer fabricated by two methods. It can be obviously seen that the distribution of seed layer prepared by the spin-coating method is more uniform than dip-coating step. Therefore, we finally use spin-coating method to prepare the seed layer.

Preparation of ZnO nanowire arrays
The growth solution was prepared with a constant concentration ratio of 1:1 between zinc nitrate and hexamethyl tetramine. To avoid scattering of precipitates onto the substrate during the growth process, the silicon wafer containing the seed layer was tilted downward at 45°into the growth solution and subsequently placed into a water bath at 95°C for growth process [15]. After the solution cooled naturally to room temperature, the sample was removed, and then thoroughly washed several times with deionized water before placed in air at 60°C for drying. The final ZnO nanowire arrays were obtained.

Results and discussion
3.1. Morphological analysis 3.1.1. Effect of seed solution concentration on the morphology of ZnO nanowires The surface morphology of the samples was characterized by scanning electron microscopy (SEM). Figure 2 shows the surface pictures of ZnO nanowires obtained with seed solution concentrations of 0.025, 0.05, and 0.075 mol l −1 , respectively. The insert shows the cross-section view. The subtracts covered with seed layer were placed in the growth solution of 0.05 mol l −1 maintained at 95°C in a thermostatically controlled bath for 3 h. It is obvious seen from figure 2 that the density of ZnO nanowires increase with the increasing concentration of seed solution. Whereas the diameter and height of nanowires change slightly. At the seed solution concentration of 0.025 mol l −1 (figure 2(a)), the orientation of the whole nanowire array was not neat enough, the gap between the nanowires were too large, and the grown nanowires had a maximum angle of 30°with the vertical direction. Further increased the seed solution concentration to 0.05 mol l −1 ( figure 2(b)), the verticality and porosity of ZnO nanowires achieve the optimal effect. When the seed solution concentration was 0.075 mol l −1 (figure 2(c)), the average diameter of the ZnO nanowires is around 100 nm and the height has not changed significantly, remaining at around 1.6 μm. Besides, the orientation of the nanowire arrays prepared at this concentration was significantly denser than that of the above two groups of samples, and the nanowires grew almost vertically to the substrate. However, the nanorod array grown at this concentration is no longer a single nanowire, partial nanowires will adhere in the growth process due to the small gap, which will reduce the porosity of the whole nanorod array. We can conclude that the seed solution mainly affects the amount of nucleation of the seed layer on the substrate and ultimately affect the density of ZnO nanowires. Whereas the seed solution has less effect on the height and diameter of the nanowires.   Figure 3 shows the SEM pictures obtained from ZnO nanowires prepared with seed solution concentration of 0.05 mol l −1 , growth solution concentrations of 0.025, 0.05, and 0.075 mol l −1 , respectively. The hydrothermal growth time was 3 h. It can be seen that the average diameter and height of the ZnO nanowires is obtained at a concentration of 0.025 mol l −1 is about 70 nm and the height of the nanowires is about 1.5 μm, and the crosssectional view shows that the orientation of the nanowires obtained at this concentration is average. When the concentration of the growth solution was 0.05 mol l −1 , the average diameter of the nanowires grew to about 100 nm and the height of the nanowires is about 2 μm. Besides, the orientation of the nanowires was greatly improved with moderate porosity. When the concentration of the growth solution was increased to 0.075 mol l −1 , a relatively dense ZnO layer with a thickness of approximately 1 μm were observed. The specific surface area of ZnO nanowires and the porosity of nanowire arrays will affect the gas adsorption. Therefore, we can conclude that the growth solution mainly controls the diameter and length of ZnO nanowires, which increases significantly with increasing concentration of the growth solution. Large specific surface area and appropriate porosity can provide more adsorption sites for gas molecules [16]. Based on the above factors, the growth solution concentration of 0.05 mol l −1 is suitable for the preparation of gas sensor.

Effect of hydrothermal growth time on the morphology of ZnO nanowires
To investigate the effect of hydrothermal growth time on ZnO nanowires, we used a seed solution concentration of 0.025 mol l −1 and a growth solution concentration of 0.05 mol l −1 for 3 h, 6 h, and 9 h, respectively. Figure 4 shows the SEM pictures of ZnO nanowires obtained with different growth time. When the growth time was 3 h, the average diameter of ZnO nanowire is about 60 nm, and the height is about 1.3 μm. When the growth time was 6 h, the average diameter of ZnO nanowire is about 90 nm, and the height is about 1.7 μm. When the growth time was 9 h, the average diameter of ZnO nanowire is about 100 nm and the height of nanowire is about 2.2 μm. We can conclude that the diameter and length of the obtained ZnO nanowires both increased with increasing growth time.

Effect of aluminium doping concentration on the morphology of ZnO nanowires
Since the gas-sensitive devices prepared from pure ZnO have many drawbacks, such as low sensitivity, slow response, and poor reliability. Thus, pure ZnO cannot meet the requirements of people's industrial production and daily life [17]. Single semiconductor oxides, such as tin dioxide, titanium dioxide, and tungsten trioxide have similar drawbacks. It is necessary to take measures to improve the gas-sensitive performance before they can be applied. Doping with impurities is a simple and reliable way to enhance the gas-sensitive properties of ZnO [18,19]. Aluminium doping concentrations of 0.5 at%, 1.0 at%, and 1.5 at% were chosen to investigate the optimal doping concentration. The seed solution concentration, growth solution concentration and growth time were 0.05 mol l −1 , 0.05 mol l −1 , and 3 h, respectively. Figure 5 shows the SEM images of the corresponding surfaces and cross sections. With the increase of the concentration of doping atoms, the growth rate of nanowires slows down, as well as the density also decreases. When the aluminium doping concentration was 0.5 at%, the diameter of ZnO nanowires is about 50 nm and the height is about 0.88 μm. The cross-sectional image shows that there is still good orientation, but the height is much shorter than that of nanowires obtained without doping under the same conditions. When the concentration of aluminium doping was 1.0 at%, the diameter of ZnO nanowires is 50 nm and the height is approximately 0.75 μm. However, the orientation is obviously deteriorated, and the grown nanowires have a maximum angle of 35°from the vertical direction. Besides, the density is also reduced. When the aluminium doping concentration was 1.5 at%, the height and density once again shrink and the orientation is poor. From the above conclusions we suspect that it is possible that the excess aluminium atoms form some kind of crystal defects that cause the density reduction and poor orientation.
The samples were subjected to x-ray diffraction (XRD) tests. Figure 6 shows XRD images of pure ZnO nanowire and ZnO nanowire doped with 0.5% aluminium atoms. As can be seen from the two curves, all diffraction peaks can be detected matching the standard XRD spectrum of ZnO with lattice constants a = 0.325 nm and c = 0.520 nm [20]. The strongest diffraction peak is in the (002) direction, which indicates the preferential growth of the sample along the c-axis direction [21]. With the increased doping of aluminium, except for the peaks at the original positions, peaks of different intensities were also detected at different positions such as (100). This result further indicates that the ZnO nanowires prepared under aluminium doping are poorly oriented and have a haphazard growth direction. Thus, the doping concentration is preferably less than 1.0 at%.

Ohmic contact properties of zinc oxide and silver
This experiment was performed by vaporizing a silver (Ag) film as an electrode and growing ZnO nanowires on the silver film to form ohmic contact. The silicon substrate was ultrasonically cleaned with acetone, isopropyl alcohol (IPA) and deionized water. After the cleaning process, the substrate was dried under nitrogen flow. The cleaned substrate was loaded into a vacuum magnetron sputter coater for bottom electrode. A 40-nm-thick Cr film was first deposited to ensure the adhesion between the silicon substrate and the silver layer. A 500-nm-thick Ag layer was subsequently deposited as the bottom electrode. It was followed by the seed layer was spin-coated on the substrate ( figure 7(a)). And then the ZnO nanowire arrays were grown by hydrothermal method with 0.05 mol l −1 seeding solution and 0.05 mol l −1 growth solution. Aluminium nitrate of 0.5 at% was added to the growth solution. After the growth process, the entire ZnO nanowire array was covered with negative photoresist (DNR-L300-D1), as shown in figure 7(d). An ohmic contact window was formed by optical lithography, and then the nanowires were developed with a developer (DPD-200) for 15 s to expose the patterned window. Figure 8 shows the AFM images of top-view ZnO nanowires with photoresist film and after developed in a developer (DPD-200) for 15 s. We can see the clear hexagonal ZnO nanowires, indicating that the zinc oxide nanowires at the ohmic window have been completely exposed. After the deposition of 400-nm-thick Ag film, the samples were immersed in acetone to remove the photoresist by lift-off process, then rinsed with deionized  water and dried in an oven at 80°C to finally form the top electrode. Finally, the samples were annealed in the rapid thermal processing (RTP) system and protected by nitrogen for 5 min Figure 9 manifests the SEM image of the cross-section of ZnO nanowire arrays on ZnO nanowires.
The ohmic contact of Ag and ZnO is studied to better realize the application of ZnO material in gas sensor devices. The samples were subjected to I-V tests with a semiconductor parameter analyser. The sample were annealed at 300°C, 400°C, 450°C, 500°C, and 600°C for 5 min. The trend of the calculated resistivity is shown in figure 10. The resistivity decreases with the temperatures from room temperature to 450°C, and then increases from 450°C to 600°C. It should be noted that the resistivity at 450°C is the lowest. Therefore, the ohmic contact is formed at 450°C. We guess that the reason for this phenomenon is that the annealing in lowtemperature N 2 in this experiment may induce diffusion between a small amount of Ag and Zn atoms, which is beneficial to improve the ohmic contact characteristic. While at high temperature annealing, a large number of atoms diffuse with each other, resulting in a large change in the stoichiometric ratio of Zn and O in ZnO, which eventually lead to the poor ohmic contact characteristic.

Conclusion
In this paper, a mixture of zinc acetate and ethanolamine was used as seed solution to fabricate ZnO seed layers on silicon substrates, and zinc nitrate and hexamethyl tetramine were used as growth solution to grow ZnO nanowire arrays by hydrothermal method. The effects of seed solution concentration, growth solution concentration, growth time and aluminium doping on the morphology of ZnO nanowire arrays were compared. After the experimental comparison, we found that the height and diameter of the hydrothermally grown ZnO  nanowire arrays increased continuously with increasing the growth solution concentration and hydrothermal time, but the change of increasing the growth solution concentration was more obvious than that of increasing the hydrothermal reaction time. The size of the seed solution concentration was proportional to the density of the ZnO nanowire arrays, and the doping impurity concentration would inhibit the growth of ZnO nanowires. Therefore, in this experiment, the seed solution concentration of 0.05 mol l −1 , the growth solution concentration of 0.05 mol l −1 , the Al doping of 0.5 at% and the growth time of 9 h were chosen as the best conditions for the preparation of ZnO nanowire arrays for gas-sensitive sensor devices.