Synthesis, Electrical Measurement, and Field Emission Properties of α-Fe2O3Nanowires

α-Fe2O3nanowires (NWs) were formed by the thermal oxidation of an iron film in air at 350 °C for 10 h. The rhombohedral structure of the α-Fe2O3NWs was grown vertically on the substrate with diameters of 8–25 nm and lengths of several hundred nm. It was found that the population density of the NWs per unit area (DNWs) can be varied by the film thickness. The thicker the iron film, the more NWs were grown. The growth mechanism of the NWs is suggested to be a combination effect of the thermal oxidation rate, defects on the film, and selective directional growth. The electrical resistivity of a single NW with a length of 800 nm and a diameter of 15 nm was measured to be 4.42 × 103 Ωcm using conductive atomic force microscopy. The field emission characteristics of the NWs were studied using a two-parallel-plate system. A low turn–on field of 3.3 V/μm and a large current density of 10−3 A/cm2(under an applied field of about 7 V/μm) can be obtained using optimal factors ofDNWsin the cathode.


Introduction
One-dimensional (1D) nanostructures have been extensively studied because of their unique chemical and physical characteristics. Because of their high aspect ratio and sharp tips that enhance the local electrical field, they can be used as emitters in field emission (FE) applications [1]. Factors that affect FE properties include the population density of the emitters (the number of emitters per unit square) [2], the electronic resistance of a single emitter [3], and the aspect ratio (radius and length) [4]. The field emission properties of an emitter are affected by nearby emitters, which is called the field screen effect. Many research groups have studied the relationship between FE properties (field screen effect) and the population density of emitters for ZnO nanowires (NWs) [5], Si NWs [6], CNTs [2,3], and CuO NWs [1]. Poor FE properties were obtained when the population density of the emitters was too high.
Investigating the electronic properties of 1D nanomaterials can reveal the relationship between the electronic properties of emitters and the FE properties of the device. In addition, it can provide useful information for the theoretical study of FE characteristics. Electronic measurement methods for NWs are generally categorized into four types: (1) laterally growing NWs between two electrodes and directly measuring them [7,8], (2) spreading the NWs on a large number of patterned electrodes and measuring a single NW on a pair of electrodes [9,10], (3) directly measuring vertical NWs using conductive atomic force microscopy (CAFM) [11,12], and (4) measuring aligned NWs arrays using two electrode films [13,14]. The CAFM system is a convenient method for the nondestructive characterization of the electronic properties of NWs.
a-Fe 2 O 3 (hematite) is a semiconductor (Eg = 2.1 eV) and the most stable iron oxide under an ambient environment [15]. a-Fe 2 O 3 NWs have recently been synthesized by various research groups [16,17]. Because they are thermally stable, resistant to oxidation, and have a high aspect ratio, a-Fe 2 O 3 NWs are a candidate emitters for FE applications. It has been reported that an electrical field of 7-8 V/lm is required to obtain a 10 -5 A/cm 2 emission current using densely packed a-Fe 2 O 3 NWs as emitters [16]. However, there have been no detailed studies on FE properties related to the population density of a-Fe 2 O 3 NWs (D NWs ) and their resistance (R NW ).
In this study, we report a simple method that can be used to control the population density of a-Fe 2 O 3 NWs (D NWs ) by varying the film thickness of iron. The FE properties of a-Fe 2 O 3 NWs with various population densities and resistances (R NW ) were studied. It was found that the best FE properties can be obtained by using the optimal D NWs in the cathode.

Experimental Details
a-Fe 2 O 3 NWs were grown using an iron film thermally oxidized at 350°C for 10 h. Iron films with thicknesses of 30 nm, 50 nm, 100 nm, and 150 nm were coated on indium tin oxide (ITO) glass by direct current (DC) sputtering. The iron-coated substrates were heated at 350°C for 10 h in an oven in the air atmosphere. The morphology and crystalline structure of the as-synthesized NWs were characterized by field emission scanning electron microscopy (FE-SEM, HITACH S-4800) and high-resolution transmission electron microscopy (HR-TEM, JEOL JSM 3010), respectively. The current-voltage (I-V) characteristics of the as-grown NWs were obtained using the CAFM (SEIKO SPA-400) system. A highly conductive polygon shaped tip with a height of 10 lm and a radius of 50 nm (the top of the tip) was used as an electronic probe while ITO was used as an electrode. The applied voltages were -10 V to ?10 V. FE properties were measured in a vacuum chamber at 6 9 10 -6 torr. The sweep voltage was 0-1100 V and the emission current was measured using Keithley 2410. Figure 1a and b shows the top view and cross-section view of the Fe-coated (50 nm) ITO glass after thermal oxidation at 350°C for 10 h, respectively. As can be seen in Fig. 1a, the NWs formed randomly on the film. Figure 1b shows that the NWs were grown vertically on the film with diameters of 8-25 nm and lengths of several hundred nm. Fe film was oxidized during the process and became a-Fe 2 O 3 film with a thickness of 100 nm. Figure 2 shows the TEM images of an a-Fe 2 O 3 NW which was grown on the iron film with a thickness of 50 nm. As can be seen in Fig. 2a, the length and diameter of the as-produced a-Fe 2 O 3 NWs were approximately 240 nm and 25 nm, respectively. Figure 2b shows the high-resolution TEM image and its corresponding selective area electron diffraction (SAED) pattern (inset). It was found that the NW is a single crystalline a-Fe 2 O 3 NW with the [110] growth direction. The interplanar spacing of 0.251 nm agrees well with the fringe spacing of a-Fe 2 O 3 . The SAED pattern reveals that the crystal structure of the a-Fe 2 O 3 NW is rhombohedral.

Results and Discussion
The length, diameter, and population density of the a-Fe 2 O 3 NWs were studied for various Fe film thicknesses. Samples A, B, C, and D represent the materials after the thermal oxidation of the 30 nm, 50 nm, 100 nm, and 150 nm-thick iron films, respectively. Figure 3a-d shows the top views of sample A (30 nm), sample B (50 nm), sample C (100 nm), and sample D (150 nm), respectively. As can be seen, the number of a-Fe 2 O 3 NWs increased with an increase of the film thickness. D NWs was calculated from the SEM image as 1.6 9 10 8 NWs/cm 2 for sample A, 7.8 9 10 8 NWs/cm 2 for sample B, 13.2 9 10 8 NWs/cm 2 for sample C, and 18.3 9 10 8 NWs/cm 2 for sample D. Table 1 summarizes the population densities of the samples. Figure 3e shows a plot of NW density (D NWs ) versus iron film thickness. The error in the D NWs measurements is about 5%. The growth mechanism of the a-Fe 2 O 3 NWs via thermal oxidation was discussed by Takagi [18] and Fu et al. [19]. They proposed that the growth mechanism of the a-Fe 2 O 3 NWs was neither vapor-liquid-solid (VLS) nor vaporsolid (VS). A tip growth mechanism was suggested, in which the NWs are grown by the diffusion of oxygen atoms on the surface defects of the iron film. During the thermal process, the oxygen molecules diffuse into the iron film between the iron atoms and form Fe-O bonds. The process involves a volume expansion in all three dimensions to accommodate the oxygen atoms. However, because the iron film prevents expansion in the two lateral directions, the volume of oxides expands upward only. A thin oxide film is therefore formed on the substrate. While a layer of the oxide forms, defects are created in the oxide film. This might be caused by a rapid thermal oxidation rate on the surface of the film during the temperature ramping period. The defects can be seen as crackings on the surface of the film (Fig. 1a). As the oxidation time increases, in regions with a higher degree of defects, the oxygen molecules move faster, diffuse across the existing oxide layer, and react with iron to form iron oxide. There was an oxidation rate difference (growth rate of a-Fe 2 O 3 NWs) between the high degree and low degree defect regions. In addition, the preferred growth orientation of a-Fe 2 O 3 NWs was the [110] direction (Fig. 2b), which suggests that the fastest growth rate was along the [110] direction. Therefore, the oxidation rate difference and the selective directional growth are probable reasons for the formation of a-Fe 2 O 3 NWs. A schematic graph of the formation of a-Fe 2 O 3 NWs is shown in Fig. 4. Samples I and II represent the thin and thick iron films, respectively. After oxidation for a period of time (stage A), the thin iron film (sample I) was completely oxidized to become a-Fe 2 O 3 NWs and a-Fe 2 O 3 film. In contrast, the thick iron film was partially oxidized so that a-Fe 2 O 3 NWs, a-Fe 2 O 3 film, and Fe film were present in stage I. At this stage, for both samples I and II, short NWs (\100 nm) and long NWs ([500 nm) grew from the low and high degree defect regions of the film, respectively. As the oxidation time increased (stage B), the length and D NWs of sample I did not change because the oxidation terminated at stage A. For sample II, the oxidation continued; NWs were continuously grown, becoming longer and more dense. Takagi [18] observed that the population density and the length of NWs were dependent of the oxidation time. Our results show good agreement with Takagi's report. The population density of NWs can be increased by increasing the thickness of the iron film.
CAFM was employed to measure the electronic property of a single a-Fe 2 O 3 NW and a-Fe 2 O 3 film. Figure 5a shows the schematic drawing of the CAFM measurement. NWs were embedded in photoresist (PR) and only the tips of the NWs (sample B) were exposed. The CAFM tip made contact with the exposed tips of the NWs so that their electronic properties could be measured. Using PR prevented the bending of the NWs while the CAFM tip approached them. The applied voltage was swept from -10 V to ?10 V to obtain the I-V characteristics for a single NW. The equivalent circuit diagram for the measurement of an a-Fe 2 O 3 NW is shown in Fig. 5b. R contact is the contact resistance between the CAFM tip and the NW. R NW and R film are the resistances of the NW and the film, respectively. The overall resistance (R overall ) is the sum of R contact , R NW , and R film . It can be expressed as: Figure 5c shows the top view and side view (inset) of a sample. As can be seen, a single NW with a diameter of 15 nm and a length of 800 nm was embedded in the PR. Only the tip of the NW was exposed on the PR surface. The figure also shows that the NWs were not bent or damaged after the spin coating process of the PR. Figure 6a and b shows the morphology image and current image, respectively, taken at a ?10 V bias on a single NW. The tip used for CAFM was an Au-coated silicon tip, with a diameter of about 50 nm. The morphology image of the NW shows that the diameter of the NW was larger than that observed by SEM, which was due to the large diameter of the scanning tip. From the current image taken at a ?10 V bias, the bright zone indicates that a current appeared on the tip of the single NW. It can also be seen that there was no current    Figure 7 shows the I-V curves of the structure of the ITO/a-Fe 2 O 3 film/a-Fe 2 O 3 NW/ CAFM-tip (Fig. 7a) and the ITO/a-Fe 2 O 3 film/ CAFM-tip (Fig. 7b). Nanowires from sample B were removed by ultrasonication in ethanol. The bias voltage was swept from -10 V to ?10 V. The measured resistances of the ITO/ a-Fe 2 O 3 film/a-Fe 2 O 3 NW/CAFM-tip structure (R overall ) and the ITO/a-Fe 2 O 3 film/CAFM-tip structure (indicated as R film ? R contact ) were 2 9 10 11 X and 4.2 9 10 9 X, respectively. Because R film ? R contact (4.2 9 10 9 X) was much smaller than R overall (2 9 10 11 X) and R contact is considered to be much smaller than R NW and R film , Eq. 1 can be rewritten as: In other words, the resistance of the ITO/a-Fe 2 O 3 film/ a-Fe 2 O 3 NW/CAFM-tip structure (R overall ) can be considered as R NW . R NW can be expressed as: where q is the electronic resistivity of an a-Fe 2 O 3 NW, and L and r represent the average length (800 nm) and radius (15 nm) of the NW, respectively. We assumed that q (4.42 9 10 3 Xcm) is a constant for all of the as-produced a-Fe 2 O 3 NWs. Using the mean length and radius of the NWs, R NW (or R overall ) of samples A, B, C, and D was calculated as 1.25 9 10 11 , 1.4 9 10 11 , 1.4 9 10 11 , and 6.25 9 10 10 X, respectively. The values are shown in Table 1.
The two-parallel-plate system used for FE measurement is shown in Fig. 8a. The cathode consisted of the a-Fe 2 O 3 Fig. 6 The CAFM images of (a) morphology and (b) current taken at a ?10 V bias emitters, a-Fe 2 O 3 film, and ITO glass while the anode was ITO glass. The cathode and anode were separated by a 150 lm-thick spacer. Figure 8b shows the curve of the field emission current density versus electrical field (J-E) for the four samples. The turn-on field (E to ) is defined as the value of the applied electrical field which produces an emission current density of 10 lA/cm 2 . To analyze the field emission properties of a-Fe 2 O 3 NWs, the Fowler-Nordheim (F-N) [20] equation was used. The equation can be expressed as: where J is the emission current density; u is the work function; F is the electrical field; A = 1.54 9 10 -6 AeVV -2 ; B = 6.83 9 10 3 eV -3/2 Vlm -1 , and b is the field enhancement factor. b is a parameter that depends on the crystal structure and the morphology of the emitters. The work function of a-Fe 2 O 3 is 5.6 eV [16], so b can be calculated as 1023, 1754, 1434, and 1420 for samples A, B, C, and D, respectively. Figure 8c shows the Fowler-Nordheim (F-N) plot of the four samples. The linear relationship of 1/E and ln(J/E 2 ) indicates that the field emission behavior of the samples fits the F-N mechanism. Table 1 summarizes the results of the turn-on field and field enhancement factor (b). Emitters with D NWs of 7.8 9 10 8 NWs/cm 2 (sample B) showed the lowest E to and the highest b in our study. Sample B had a medium population density of the a-Fe 2 O 3 emitters.
In contrast, emitters with low D NWs (sample A) showed the lowest b and a ''modest'' E to due to insufficient emission sites. [3] Densely packed NWs (samples C and D) caused the screen effect [21], leading to field emission performance that was not as good as that of sample B. Compared to other 1D materials, such as ZnO NWs, [5] Si NWs, [6] CNTs [2,3], and CuO NWs, [1] NWs with high b and low E to can be achieved by controlling their population density. Our study agrees well with other studies. The relationship between the field enhancement factor (b), resistance of a-Fe 2 O 3 NWs (R NW ), and the population density (D NWs ) of the four samples was analyzed; it is shown in Fig. 9. R NW of samples A, B, C, and D are similar, but b is different for each sample. According to the results, the value of b depends on D NW more than it does on R NW . Bonard et al. [21] observed that b depends on D NW more than it does on the length of CNT. The length of CNT increased with an increase in resistance. Our results show good agreement with Bonard's research.

Conclusion
The formation, electronic characterization, and field emission application of a-Fe 2 O 3 NWs were studied. a-Fe 2 O 3 NWs were grown vertically on the substrate via the thermal oxidation of an iron film in air at 350°C for 10 h. By increasing the film thickness (to 30, 50, 100, and 150 nm), D NWs could be increased. R NW of the NWs was estimated using the CAFM technique. In the FE study, an E to of 3.3 V/lm and a large current density of 10 -3 A/cm 2 (under an applied field of about 7 V/lm) were obtained with the optimum D NWs . The densely packed NWs caused the screen effect, leading to poor FE performance. The field enhancement factor (b) depends on D NW more than it does on R NW . This study shows that a-Fe 2 O 3 NWs are a candidate for emitters in field emission applications.