Mapping of Work Function in Self-Assembled V₂O₅ Nanonet Structures

Transition metal oxides (TMO) including vanadium pentoxide (V₂O₅) have been widely used for the modern organicbased electronic devices such as organic light emitting diodes (OLED), organic photovoltaic (OPV) cells, field emission devices and photovoltaics due to their high quality electrical and optical properties.¹⁻⁷ For example, thermally deposited V₂O₅ thin film has recently been reported as charge transport layer in high-performance polymer organic field effect transistor (OFET).⁸ In this case, the formation of V₂O₅ layer exhibited charge carrier mobility comparable to that of a pristine Au electrode, enabling the large-area and low-cost electronics. Another example is the three-dimensional structures of V₂O₅ nanosheets as electrodes in supercapacitors, which can be utilized for energy storage.⁹ Although V₂O₅ have the exceptional properties, the surface potential measurements of V₂O₅ nanostructures remain challenging and have yet to be characterized experimentally in a nanoscale. Scanning Kelvin probe microscopy (SKPM) is very powerful tool to measure the contact potential difference (ΔV_cpd) between a metal tip and a sample surface for analyzing the local chemical potential distribution, surface charging, dopant concentration and work function of the various nanostructures.¹⁰⁻¹⁵ Here, we measured the work function of the V₂O₅ nanonet structures on SiO₂ substrate by using SKPM techniques in ambient conditions, where the V₂O₅ nanonet structures were self-assembled through dropping the V₂O₅ nanowire (NW) solution and dried the solvent on the substrate. SKPM based on an atomic force microscopy (AFM) maps the ΔV_cpd between the metal tip and the V₂O₅ nanonet. By generating the histograms of the ΔV_cpd obtained from the SKPM maps for the V₂O₅ nanonet as well as the HOPG as a reference for a calibration of the SKPM tip, we successfully estimated the work function of ~5.1 eV for the V₂O₅ nanonet structures.

V₂O₅ NW solution was prepared by mixing 1 mg ammonium (meta) vanadate (Aldrich) with 10 mg acidic ionexchange resin (DOWEX 50WX8-100, Aldrich) in 20 mL water as previous processes.^16,17 V₂O₅ NWs with lengths of a few micrometers were synthesized after keeping the solution for about 10−20 days. We dropped the V₂O₅ NW solution on the substrate and evaporated the solvent, leaving the V₂O₅ nanonet on the substrate which was a p-type degenerately doped Si wafer with 100 nm silicon oxide layer as shown in Fig. 1(a). Fig. 1(b) shows a schematic diagram of the SKPM experiments of the assembled V₂O₅ nanonet on the substrate. The SKPM measurements were performed by a commercial AFM based on a Park XE7 (Park Systems, Korea) in ambient conditions at room temperature. We used a Cr/Au coated tips (NCSTAu, Nanosensors) with a spring constant (k) of 7.4 N/m and a resonance frequency (ω₀) of 160 kHz to obtain the topography and ΔV_cpd images simultaneously in a non-contact mode. Electrical contacts to the substrate were grounded during the measurements.

Figure1.

(a) Dropping the V₂O₅ NW solution on SiO₂ substrate. (b) Schematic diagram for SKPM setup. The tip scans over the V₂O₅ nanonet while oscillating at the resonance frequency of ω₀, in addition to the AC and the DC bias with the external lock-in amplifier which separates the ω frequency from the output signals. The surface potential is measured by reading the ω signal and the DC bias control in the feedback loop.

We applied an AC modulation voltage (V_acsin(ω t)) and a DC bias (V_dc) voltage between the metal tip and the V₂O₅ nanonet, and measured the ΔV_cpd by using an external lock-in amplifier (SR850, Stanford Research Systems) to analyze the amplitude of the tip oscillation. The tip oscillates due to the electrical force between the metal tip and the sample surface, where the ω component is separated from a whole output signal by the external lock-in amplifier. The total bias voltage between the tip and the V₂O₅ nanonet is

If we assume the tip-sample system as a parallel plate capacitor, the electrostatic force between the tip and the sample is given by:

where C is a tip-sample capacitance and z is a tip-sample distance. The electrostatic force at a frequency ω, has the following expression

This force is zero when the V_dc equals to the ΔV_cpd, so the DV_cpd can be measured by applying V_dc such that the oscillating amplitude at the frequency of ω is nullified by using the external lock-in amplifier that measures the w signal is zero in feedback loop.^11−13,18 In this case, we applied an AC modulation voltage amplitude (V_ac) of 1.5 V at a frequency (ω) of 17 kHz to the Au-coated tip.

Fig. 2(a) shows the transmission electron microscope (TEM) images of V₂O₅ nanonet structures. The individual V₂O₅ NWs have the diameters of ~10−20 nm and the lengths of ~over 1 µm long as shown in previous results.^16,17 The V₂O₅ nanonet is randomly composed of the individual, bundles and stacked NWs. Fig. 2(b) shows the AFM topography image of the V₂O₅ nanonet on SiO₂ substrate. The thick NWs in the V₂O₅ nanonet indicate the bundle or stacked NWs after evaporation of the NW solution on SiO₂ substrate.

Fig. 3(a) is the expanded AFM topography image with various thicknesses of the V₂O₅ nanonet, corresponding to the white dashed square in Fig. 2(b). Fig. 3(b) is SKPM image obtained simultaneously with the topography image by measuring the local ΔV_cpd between metal tip and the V₂O₅ nanonet. The bottom graphs show the thickness (left) and the ΔV_cpd (right) profiles along the marked regions in each AFM topography and SKPM image. The entire surface of cantilever (cone) which is coated with metal film also can be effected by the chemical potential from the sample, however only a few atoms at the end of the probe interact with the sample through attractive or repulsive forces. Therefore, the resolution of the topography image is higher than that of the SKPM image as shown in Fig. 3. In the SKPM image, the brightness variations come from the ΔV_cpd between the Au-coated tip and the sample surface. The bright region presents the smaller ΔV_cpd with respect to the Au-coated tip than the dark region, indicating smaller work function of V₂O₅ nanonet compare to that of the SiO₂.¹⁹

Figure2.

(a) TEM images of the V₂O₅ nanonet. (b) AFM topography image of the self-assembled V₂O₅ nanonet on SiO₂ substrate.

Figure3.

(a) and (b) AFM topography and corresponding SKPM images in the region of the white dashed square in Fig. 2(b). The thickness (gray) and the ΔV_cpd (red) profiles along the marked regions in AFM topography and SKPM image are shown in the bottom. Inset shows the statistical histogram for the ΔV_cpd in the region of V₂O₅ nanonet.

We calibrate the metal Au-coated tip with reference sample of HOPG to determine the work function of the SKPM tip (f_tip) that has been used in the experiments. The cantilever amplitude of the first harmonic (ω) signal is zero when the DC offset bias (V_dc) is equal to the surface potential of the sample surface. Fig. 4(a) shows the electric force microscopy (EFM) cantilever amplitude as a function of the tip-sample bias voltage relative to the HOPG substrate (gray) and the V₂O₅ nanonet (brown). As shown in the graph, the amplitude as the electrical force can be nullified if an applied external bias (V_dc) is equal to the ΔV_cpd that corresponds to the work function difference between the Autip and the sample.¹³ The DC offset is about 0.15 V for the V₂O₅ nanonet (brown) and about −0.3 V for the HOPG surface (gray) as shown in the graph. Therefore, we could also obtain the work functions of the Au-tip and the V₂O₅ nanonet surface by using the two curves. The ΔV_cpd between the Au-coated tip and the HOPG is about −0.3 V, so the work function of the Au-tip is determined to be 4.95 eV from the work function of the HOPG (f_HOPG = 4.65 eV) as a reference. Therefore, we can estimate the work function value (f_s) of the V₂O₅ nanonet surface by using the relation¹³

and it is found to be about 5.1 eV, which is almost same to the work function of the solution processed V₂O₅ film layer (~5.3 eV)^20,21 and a bit smaller value than the that of the V₂O₅ film layer, ~6.8 eV, which is fabricated in the ultra-high vacuum, results from the ultra-violet photoemission spectroscopy (UPS) and inverse photoemission spectroscopy (IPES) measurements.^5,22,23 The decreasing work function probably arise from the air exposure resulting in the shift of the photoemission onset towards lower binding energy due to an additional V⁴⁺ surface state, which is equivalent to a downwards shift of the vacuum level.²⁰ Fig. 4(b) shows the representative band diagram, depicting the work function values of the HOPG (f_HOPG), the Au tip (f_tip), and the V₂O₅ nanonet (f_s). It indicates that the V₂O₅ nanonet has higher work function than those of the HOPG and the Au tip.

Figure4.

(a) EFM cantilever amplitude versus tip-sample bias voltage plots for the HOPG (gray) and the V₂O₅ nanonet surface (brown). (b) Representative band diagram for work functions of the HOPG, the Au tip, and the V₂O₅ nanonet. The black dashed line indicates the vacuum level energy.

In conclusion, we investigated the surface potential of the V₂O₅ nanonet by using SKPM technique in ambient conditions. We successfully obtained the work function of ~5.1 eV for the V₂O₅ nanonet by calibrating the Au-tip with reference sample of HOPG. This value is almost same to the work function of the solution processed V₂O₅ film layer measured by the UPS and IPES in the literature. This SKPM measurement allows us to evaluate the electronic properties of the V₂O₅ NW based devices in future.