Synthesis of vanadium pentoxide nanoneedles by physical vapour deposition and their highly sensitive behavior towards acetone at room temperature

Abstract

V₂O₅ nanoneedles were synthesized using a facile physical vapour deposition approach. The XRD patterns confirmed the good crystallinity of the as prepared nanoneedles. The chemical states of the V₂O₅ nanoneedles were confirmed through XPS analysis. Gas sensor based on the V₂O₅ nanoneedles was investigated for four types of gases (acetone, ammonia, ethanol and propylamine) at room temperature. A high reproducible response and a good selective behavior towards acetone were observed in both low and high concentration zones with a low detection limit of 941 ppb, which was explained on the basis of the energy band model. A sensing mechanism has also been suggested.

---

1. Introduction

In order to support the community, the increasing implementation of nanotechnology in different applications, such as supercapacitors, catalysis, solar cells and chemical sensors etc., has launched a competitive research run among materials scientists. In particular, scientists have concentrated on chemical sensors for their distinct services including laboratory tests, environmental control, food and other industries, alcohol detection and security purposes. Metal oxide nanomaterials with various kinds of morphologies have been explored for the detection of different volatile organic compounds (VOCs). These multi-dimensional nanomaterials exhibit an increasing number of active sites and hence show very good sensing properties.¹ A perfect sensor must have more characteristics such as the ability to detect low analyte concentrations, stability for long term usage, a fast response time, good selectivity in the presence of interfering analytes and a reasonable working temperature to avoid high power consumption.²

Among VOCs, acetone is more volatile; it is often used as an organic solvent or chemical intermediate in the laboratory and in industry. Recently, some reports have disclosed certain side effects of acetone on human health whether it is produced by the body itself (internally) or inhaled from the external environment. These include irritation of the eye and respiratory system, biliousness, indolence and drowsiness of the nervous system upon continuous exposure to acetone.³ Further, the emission of acetone during the long term storage of potatoes⁴ also conveys the significance of its detection in food industries. Moreover, the need of acetone gas sensors as breath markers to type-I diabetes is also pressing as highlighted by certain reports.^5–7

Various sensing materials have been explored aiming at the detection of acetone, such as TiO₂ thin films¹ and nanoparticles,⁸ Nb-doped TiO₂,^9,10 Au-doped ZnO nanowires,¹¹ etc., however, the detection limit and working temperature were too high, limiting their application to food storage and other industries, which results in huge power consumption. Selectivity is also a widely accepted issue for gas sensors; to date, there are few works focusing on the high selectivity towards acetone based on the above nanomaterials.

Vanadium pentoxide (V₂O₅) is an n-type transition metal oxide with an interesting layered structure and an electronic conductivity of ∼0.5 S cm⁻¹ at room temperature.¹² It is a non-stoichiometric material which is known for its catalytic properties in oxidation reactions.¹³ These properties make it well suited for the construction of functional materials and novel devices for operation under ambient conditions. One-dimensional V₂O₅ nanostructures (nanorods, nanobelts, and nanotubes etc.) have already been investigated for the sensing of different gases such as ethanol, ammonia, amine and toluene.^12,14–16 However, the major drawback with these nanostructures is their functional limitation to detect only high concentrations of these gases. Their acetone sensing behaviour has also not been studied to date.

In this work, we report the synthesis of V₂O₅ nanoneedles using a physical vapour deposition approach in a controlled atmosphere. Gas sensor based on the V₂O₅ nanoneedles exhibits significant response towards acetone at room temperature even in a very low concentration range with a reasonable response time. The high selectivity of the sensor and the stable response towards acetone at low concentration levels are explained on the basis of the energy band model, and a reasonable mechanism has also been suggested.

2. Experimental

2.1 Synthesis of V₂O₅ nanoneedles

V₂O₅ nanoneedles were grown using the following procedure. Si (001) substrates, polished with silicate on one side, were pre-treated by sonication in ethanol/deionized water for 10 min. After drying in a conventional oven at 80 °C under air, the substrates were loaded on a quartz boat filled with 500 mg V₂O₅ powder (purity: 99%, Shanghai Shan Pu Chemical Engineering) as a raw source material and the silicate surface of the substrate faced down towards the source powder at a vertical distance of 5 mm. The quartz boat was then inserted inside a glass tube (one meter in length) exactly at the mid-point, which was placed inside a horizontal tube furnace under air and sealed tightly. The temperature was increased to 450 °C (growth temperature) at a rate of 10 °C min⁻¹ under air and subsequently from 450 to 650 °C at a rate of 0.5 °C min⁻¹ to stabilize the growth. At this stage, Ar gas was injected with a flow rate of 0.2 L min⁻¹ and the heating was continued towards 710 °C, beyond the melting point of V₂O₅ (∼690 °C), at a rate of 3 °C min⁻¹. The quartz tube was maintained at 710 °C for 40 min under an Ar flow of 0.1 L min⁻¹ and was finally cooled to room temperature such that the flow of Ar gas was stopped at the temperature value of 690 °C. It was noted that after deposition, the silicate surface of the substrate turned yellowish.

2.2 Characterization

The deposited nanoneedle products were then characterized using X-ray diffraction (XRD, D/MAX-III, Cu Kα) and high magnification scanning electron microscopy (SEM, Zeiss Ultra Plus, ZEISS, Germany) for investigating the structure and morphology, respectively. X-ray photoelectron spectroscopy (XPS) measurement was performed with an Escalabmk-II XPS apparatus (VG Scientific, England) with an Al target. The emission angle between the photoelectron beam and the sample surface was 45°, and the binding energy of the electron spectrometer was calibrated using the maximum adventitious C1s signal at 284.6 eV with the solved full width at half maximum (FWHM) being 0.8 eV.

2.3 Fabrication and measurements of the gas sensor

For the gas sensing measurements, the deposited V₂O₅ nanoneedles were scraped off of the substrate surface in the following manner. Substrate with the grown V₂O₅ nanoneedles was heated at 60 °C under air, followed by sonication in ethanol for 30 min. Ethanol was evaporated and the nanoneedles were mixed with terpineol to obtain a smooth paste. The paste was then coated on the surface of a ceramic tube, on which a pair of Au electrodes was already printed. This device was dried at 100 °C for 3 days in air to decompose the terpineol completely. Finally, a Ni–Cr heating wire was inserted into the tube. Before testing the gas sensing properties, the fabricated V₂O₅ nanoneedle sensor was aged at 60 °C for 5 days to improve the stability of the sensitive materials.

The gas sensing properties were tested on a WS-30A measuring system containing a static gas distribution chamber with a volume of 18 litres. During the testing process, the gas sensor was placed in the centre of the test chamber, in serial connection with an external resistor named R_L. In the gas testing process, the liquid vaporization method was used to produce the target gas with a certain concentration. Typically, a certain amount of liquid was injected into a heater base fitted inside the test chamber by means of a micro-syringe. When heated, the liquid got volatilized into the corresponding test gas, which forms a simulated environment of indoor air with gaseous vapours. Four types of liquid analytes (acetone, ammonia, ethanol and propylamine) were taken for tests at room temperature. A relation between the volume of the injected liquid and the concerned gas concentration can be derived from a general formula,¹⁷ and is written as,

where Q denotes the injected liquid volume in μL, M is the molar mass of the test gas, d is the liquid density, p is the corresponding liquid concentration and C is the gaseous concentration in ppm. The response was defined as the ratio of the resistance in air (R_a) to that in the target gas (R_g), namely S = R_a/R_g.

3. Results and discussion

XRD pattern of the as-prepared sample is shown in Fig. 1(a) and the diffraction peaks are indexed to orthorhombic V₂O₅ (JCPDS 01-086-2248). Besides, two peaks at 2θ = 10.60° and 12.20° also appear in the pattern which are related to the (021) and (211) planes of the SiO₂ phase of the Si substrate and no other impurities are observed which indicates the good crystallization degree of the V₂O₅ nanoneedles.

Fig. 1 (a) XRD pattern of the V₂O₅ nanoneedles deposited on the Si substrate, (b) the low magnification and (c and d) high magnification SEM images of the V₂O₅ nanoneedles, (e) the TEM image and (f) HRTEM image of a V₂O₅ nanoneedle with surface details (SAED pattern as the HRTEM image inset).

SEM images in different magnifications of the V₂O₅ products are shown in Fig. 1(b–d). The low magnification image (Fig. 1(b)) reveals the uniform growth of V₂O₅ nanoneedles on the Si substrate. In order to obtain a further clear insight into the microstructure, high magnification images are displayed in Fig. 1(c and d). It can be seen that each nanoneedle consists of a wide base and a sharp tip. Typical diameters of the base part and the sharp tip are in the ranges of 30–40 and 11–15 nm, respectively, and the length of the nanoneedle is 100–200 nm. Such a conical-shaped nanoneedle might be effective for better gas adsorption and electron transport during interaction with the analyte molecules because they form disordered chains in the form of a zigzag geometry when they are randomly oriented over the ceramic tube in the gas sensing device. Such a network of nanoneedles is responsible for frequent gas adsorption.^18–20

Fig. 1(e and f) display TEM and HRTEM images of the V₂O₅ nanoneedles with the corresponding SAED pattern (inset). The images clearly indicate the preferential growth of the nanoneedles along the (010) plane, which has a significant effect on the analyte-surface adsorption.²¹ The effect will be discussed in detail in the sensing mechanism part.

X-ray photoelectron spectroscopy (XPS) data for a V₂O₅ nanoneedle sample is shown in Fig. 2. The survey spectrum in Fig. 2(a) indicates the appearance of Si2p, C1s, V2p and O1s peaks. The Si2p peak appears as the nanoneedles were deposited on a Si substrate and the C1s peak appears due to the electrodes during the test. Fig. 2(b) displays the fitted profile of the vanadium peak, which shows that the V2p^3/2 peak appears at a binding energy of 517 eV, related to the V⁵⁺ state of V₂O₅ phase.^22–25 Si2p appears at a binding energy of 103.53 eV and corresponds to the Si⁴⁺ state of the SiO₂ surface of the Si substrate and is shown in Fig. 2(c). The oxygen peaks lie at three binding energies 530.47, 532.52 and 533.55 eV, related to the 1s state of oxygen for V₂O₅, SiO₂ and the adsorbed oxygen ions, respectively (Fig. 2(d)).^26,27

Fig. 2 XPS spectrum of the V₂O₅ nanoneedles deposited on the SiO₂ substrate: (a) survey spectrum, (b) vanadium 2p^3/2 peak, (c) silicon 2p peak, (d) oxygen 1s peaks.

The resistance change of the V₂O₅ nanoneedle sensor towards acetone was measured as shown in Fig. 3(a), in which the variation of the resistance in steady state mode is plotted. Acetone was injected with two concentration zones, the low concentration zone is comprised of a range from 863 ppb to 4.3 ppm and the high concentration zone ranges from 8.63 to 140 ppm. The resistance of the V₂O₅ nanoneedle sensor decreases upon the injection of the acetone gas and reaches its baseline value when the gas is released. The decline in resistance amplifies with the increasing acetone concentration, which is characteristic of the n-type V₂O₅ sensor. The corresponding sensitivity response is shown in Fig. 3(b), the sensitivity for 1.7 ppm acetone is low (S = 1.025) and it increases monotonically in the low concentration zone. A sharp upturn in sensitivity is noticed after the injection of acetone in the high concentration zone, which obviously proves the significance of the sensor for higher concentrations and its usefulness in the food industry, in particular to protect the mass spoilage of potatoes at room temperature.⁴

Fig. 3 (a) Steady state resistance transient time curve of the V₂O₅ nanoneedle-based sensor for acetone at room temperature; (b) sensitivity response versus time for different acetone concentrations; variation of % sensitivity response with acetone concentrations in the low (c) and (d) high zones.

Fig. 3(c and d) describe the variation of the sensitivity response with the acetone concentration and it is noted that the increase in sensitivity with the concentration follows an almost similar curvature pattern in both the low and high concentration zones, which shows its credibility of being feasible for higher concentrations beyond the selected zones, by symmetry. The detection limit for acetone is estimated to be 941 ppb using eqn (2),

DL = KS_o/m (2)

where K is the signal to noise ratio and it is a numerical factor chosen according to the desired confidence level and generally equal to 3. S_o is the standard deviation of the blank measurements (n = 120) and m is the slope of the response versus concentration curve throughout the range of concentrations.²⁸

The real value response behaviour of the sensor was also studied to estimate the response times at an extremely low concentration (1.7 ppm) (Fig. 4(a)). A response time of 73 s is noted, which decreases slightly with the increase in concentration till it verges on 67 s for 140 ppm. Fig. 4(b) presents the real value resistance variation of the sensor towards all the four gases, which predicts the response times of 67 s, 49 s, 34 s and 88 s towards 140 ppm of acetone, ethanol, ammonia and propylamine, respectively. After releasing the gas, the sensor resistance recovers quickly to the base-line state. The variation of the response time with the acetone concentration in both the low and high zones is plotted in Fig. 4(c and d), which shows a minute decrease in response time with increasing acetone concentration in both zones, thus representing the steady response behaviour of the sensor.

Fig. 4 (a) Real value response curve of the V₂O₅ nanoneedle sensor towards 1.7 ppm acetone to estimate the response time, (b) real value response curves of the V₂O₅ nanoneedle sensor towards 140 ppm of various gases to estimate the response times, (c and d) variation of the response time in the low and high concentration zones, respectively.

The reproducibility of the gas response was investigated by switching the acetone gas on and off at the detection threshold level (1.7 ppm). Fig. 5 shows the repetitive response of the V₂O₅ nanoneedle sensor towards acetone at room temperature. Based on the results of four repeated steps of acetone gas modulation, the reproducibility of the V₂O₅ acetone gas sensor is observed by calculating the coefficient of variation (CV), which is found to be 0.0211 (2.11%) as defined by CV = σ/μ, where σ is the standard deviation and μ is the mean.²⁹ This observation indicates that the V₂O₅ nanoneedles have stable and good sensing characteristics for acetone gas. It also shows that such sensors show good recovery to the base-line with some signal distortion.

Fig. 5 Response curve of the V₂O₅ nanoneedle sensor towards 1.7 ppm acetone over four repeated cycles of exposure.

To predict the selectivity, the steady state response of the sensor based on V₂O₅ nanoneedles was obtained in accordance with changing concentrations of different gases (acetone, ammonia, ethanol and propylamine) at room temperature. The sensor did not respond to gases other than acetone in the low concentration zone, which implies that it is perfectly selective at low concentrations. In the high concentration zone (8.63–140 ppm), the sensor detected all the four gases, however, the response was more prominent towards acetone such that a higher sensitivity (∼2.37 for 140 ppm) towards acetone vapour was achieved (Fig. 6(a)). In order to evaluate diversity in response, gas sensor was exposed to successive injection of the entire four gases one after the other, and its response to all these gases was noted as shown in Fig. 6(b). It was figured out that the sensor significantly respond towards ethanol in a stable manner than the rest of the gases. The selectivity of the V₂O₅ nanoneedle sensor in the high concentration zone was evaluated via the % response magnitudes corresponding to 200 ppm of acetone, ammonia, ethanol and propylamine, which turned out to be 248, 162, 142 and 135%, respectively (Fig. 6(c)). This obviously indicates the acetone selective nature of the V₂O₅ nanoneedles in the high concentration zone as well. Comparison of the response properties of the sensor based on V₂O₅ in this work with those of other kinds of V₂O₅ nanostructure sensors reported earlier is described in Table 1.

Fig. 6 (a) Steady state response curves of the V₂O₅ nanoneedle sensor to acetone, ethanol, ammonia and propylamine in the high concentration zone, (b) real value resistance continuum to all gases in dynamic mode, (c) selectivity in the high concentration zone, (d) linear fit calculation for estimating oxygen ion adsorption.

Table 1 Comparison of the sensing properties of the V₂O₅ nanoneedles synthesized in this work with other V₂O₅ nanostructures reported in the literature

Material	Detection threshold (ppm)	Magnitude of sensitivity	Operating temperature (°C)	Selectivity	Reference
a RT, room temperature.
V2O5 nanoneedles	1.7	2.37 for 140 ppm	RT^a	Acetone	This work
V2O5 nanorods	100	1.04 for 500 ppm	RT^a	Ethanol	12
V2O5 nanobelts	5	3 for 1000 ppm	200	Ethanol	14
V2O5 nanotubes	—	1.8 for 1000 ppm	230	Ethanol	30
Fe2O3-activated V2O5 nanotubes	10	2.1 for 1000 ppm	230	Ethanol	30

---

The sensing mechanism of the V₂O₅ nanoneedle sensor may be explained on the basis of the energy band model.¹⁷ We know that V₂O₅ is an n-type semiconductor with intrinsic defects due to stoichiometric deviation in its crystal, such as oxygen vacancies.

Quantitatively, the conductivity change is governed by the E_g value which is defined as E_g = E_c − E_v for n-type semiconductors and can be derived as follows.

The carrier density in semiconductors is expressed as,

where n_o and p_o represent the electron and hole densities, which are the majority of charge carriers for n-type and p-type semiconductors, respectively. m^*_n(p) is the effective mass of the electrons or holes. k_B and h are Boltzmann and Plank’s constants. E_f represents the Fermi level and E_cv is the highest band which is the conduction band for n-type and the valance band for p-type semiconductors and here we consider it as E_c of the conduction band.³¹ The conductance (conductivity) is directly proportional to the carrier concentration, which is equal to n_op_o and can be written as,where, at a certain temperature, m^*_n and m^*_p are constants, so the conductivity is mainly controlled by the exponential term (exp(−E_g/(k_BT))) or simply by the E_g value, which is related to the adsorption of analyte molecules to the sensor surface in following manner.

The adsorption of analyte (acetone) molecules on the V₂O₅ nanoneedle surface occurs in two steps including physical adsorption and chemical adsorption. Oxygen molecules from the air are physically adsorbed on the nanoneedle surface and trap electrons in them to form chemically adsorbed oxygen ions through the following reaction.

O_2(ads) + e⁻ → O₂⁻ (5)

Such adsorption of molecular oxygen ions is confirmed by the linear fit of log(S_g − 1) = log C (Fig. 6(d)). The slopes of the linear fit equations corresponding to acetone, ethanol, ammonia and propylamine turn out to be 0.32, 0.34, 0.45 and 0.37, respectively, which are close to 0.5 (the typical value for oxide ion adsorption),^32,33 and therefore predict the adsorption of such ions on the nanoneedle surface at room temperature. However, this is an exothermic reaction and the chemisorption rate is slow at room temperature,¹⁴ which is the phenomenological reason for the relatively low sensitivity value of the nanoneedle sensor. These oxygen ions distribute themselves on the V₂O₅ surface in the form of VO surface groups, which are responsible for the catalytic oxidation of hydrocarbons. In V₂O₅, the VO groups are located on the (010) plane because this surface is the lowest free energy surface in V₂O₅.²¹ This produces an electron depletion region of high electric potential (qV_s) due to the high baseline resistance of the sensor.

After the injection of analyte (e.g. acetone), its molecules are chemisorbed on the V₂O₅ surface. The acetone molecule tends to be adsorbed on the stable and lowest free energy surface (010) of the V₂O₅ lattice, which is shown schematically in Fig. 7. The reducing acetone molecule will react with the chemisorbed oxygen ion O₂⁻ and liberate the trapped electron back to the V₂O₅ surface. The ideal chemical reaction is as follows.

(CH₃)₂CO + 4O₂⁻ → 3CO₂ + 3H₂O + 4e⁻ (6)

Fig. 7 Schematic sensing mechanism of the V₂O₅ nanoneedles towards acetone.

The detailed oxidation mechanism of acetone involves the formation of intermediate products, acetate and formaldehyde, on the V₂O₅ surface, and then ultimately produces the final product CO₂.³⁴ Acetate reacts with a hydrogen atom to form a nearby hydroxyl group and is stabilized as acetic acid. This step is exothermic by 0.6 kcal mol⁻¹. Similarly, the oxidation of intermediate formaldehyde on the (010) surface of V₂O₅ is also exothermic by 22.1 kcal mol⁻¹.³⁵ Besides, the final reaction involving the production of CO₂ is exothermic by 30 kcal mol⁻¹. Such exothermic reactions are responsible for the self-evaporation of H₂O which is the final by-product in the acetone oxidation. However, in order to avoid its effect on the stability of the sensor surface, extreme care was taken during recording of the gas sensing measurements, e.g. prior to subsequent measurements, a little bit of heat treatment was applied to the sensor, followed by a certain time delay for letting the sensor surface cool down to room temperature.

The oxidation reaction leads to lowering of the qV_s value and hence reduces the electron depletion layer, which further decreases the resistance of V₂O₅. The adsorption leads to charge transfer from the acetone molecule to the V₂O₅ nanoneedle surface through the chemisorbed oxygen ions and the adsorption induced reconstruction of the V₂O₅ surface. These two mechanisms are responsible for the surface charge distribution on the V₂O₅ nanoneedle and therefore a change in the band gap energy E_g is observed, which produces a change in the conductance of the sensor.

The higher selectivity of the V₂O₅ nanoneedle sensor to acetone over other analytes may be explained on the basis of the chemical reactivity of the reducing gas molecules at the sensor surface. Acetone contains a carbonyl functional group. Due to the greater electronegativity of oxygen, the carbonyl group is a polar functional group and therefore acetone has a larger dipole moment (D = 2.88).³⁶ The oxygen atom in the carbonyl group of acetone has a higher electron density due to the lone pairs than that of the carbon atom. The reactivity of acetone may be rationalized taking into account the important resonance contributor comprising a positive carbon and a negative oxygen atom. the polarity of the group has a profound effect on the chemical reactivity of acetone. The most energetically favourable reaction pathway is that a surface oxygen atom attacks the carbonyl carbon to form a C–O bond with a bond length of 1.58 Å during the oxidation of acetone, while the activated C–C bond is elongated to 1.93 Å. By overcoming an energy barrier of 31.5 kcal mol⁻¹, the breaking of the C–C bond leads to an acetic group, CH₃COO* (acetic acid), and a formaldehyde molecule (O* denotes the lattice oxygen) on the surface. Such a favourable deep oxidation of acetone over the V₂O₅ (010) surface plane has already been confirmed using a periodic DFT method,^35,37 which shows good selective nature of gas sensors based on a V₂O₅ surface.

4. Conclusions

V₂O₅ nanoneedles were synthesized through an effective physical vapour deposition method. Gas sensor based on V₂O₅ nanoneedles exhibited significant sensing properties towards acetone compared to those towards ammonia, ethanol and propylamine at room temperature with an almost steady response time, particularly in the high detection zone (8.63–140 ppm). The tremendous sensing properties have been explained qualitatively as well as quantitatively, which implies its potential application for practical purposes specifically in food industries. A suitable sensing mechanism has also been suggested for the V₂O₅ nanoneedle sensor.

Acknowledgements

This work is supported by the International S&T Cooperation program of China (ISTCP) (no. 2013DFR50710), Equipment pre-research project (no. 625010402), Science and Technology Support Program of Hubei Province (no. 2014BAA096), the National Nature Science Foundation of Hubei Province (no. 2014CFB165), and Ministry of Education and Science of Russia (no. 14.613.21.0002).

References

1. S. Capone, A. Forleo, L. Francioso, R. Rella, P. Siciliano, J. Spadavecchia, D. S. Presicce and A. M. Taurino, J. Optoelectron. Adv. Mater., 2003, 5, 1335.
2. Z. Liu, T. Yamazaki, Y. Shen, T. Kikuta, N. Nakatani and T. Kawabata, Appl. Phys. Lett., 2007, 90, 1731191.
3. B. Bhowmik, K. Dutta, A. Hazra and P. Bhattacharyya, Solid-State Electron., 2014, 99, 84.
4. B. P. Costello, R. J. Ewen, H. E. Gunson, N. M. Ratcliffe and P. T. Phillips, Meas. Sci. Technol., 2000, 11, 1685.
5. A. M. Diskin, P. Spanel and D. Smith, Physiol. Meas., 2003, 24, 107.
6. C. H. Deng, J. Zhang, X. F. Yu, W. Zhang and X. M. Zhang, J. Chromatogr. B: Anal. Technol. Biomed. Life Sci., 2004, 810, 269.
7. M. Righettoni, A. Tricoli and S. E. Pratsinis, Anal. Chem., 2010, 82, 3581.
8. A. Teleki, S. E. Pratsinis, K. Kalyanasundaram and P. I. Gouma, Sens. Actuators, B, 2006, 119, 683.
9. V. Galstyan, E. Comini, G. Faglia, A. Vomiero, L. Borgese, E. Bontempi and G. Sberveglieri, Nanotechnology, 2012, 23, 2357061.
10. S. Phanichphant, C. Liewhiran, K. Wetchakun, A. Wisitsoraat and A. Tuantranont, Sensors, 2011, 11, 472.
11. S. J. Chang, T. J. Hsueh, I. C. Chen, S. F. Hsieh, S. P. Chang, C. L. Hsu, Y. R. Lin and B. R. Huang, IEEE Trans. Nanotechnol., 2008, 7, 754.
12. A. D. Raj, T. Pazhanivel, P. S. Kumar, D. Mangalaraj, D. Nataraj and N. Ponpandian, Curr. Appl. Phys., 2010, 10, 531.
13. C. V. Ramana, O. M. Hussain, N. B. Srinivasalu, C. Julien and M. Balkanski, Mater. Sci. Eng., B, 1998, 52, 32.
14. J. Liu, X. Wang, Q. Peng and Y. D. Li, Adv. Mater., 2005, 17, 764.
15. I. Raible, M. Burghard, U. Schlecht, A. Yasuda and T. Vossmeyer, Sens. Actuators, B, 2005, 106, 730.
16. S. A. Hakim, Y. Liu, Y. Lu and W. Chen, Mater. Sci. Semicond. Process., 2015, 31, 630.
17. W. Tang and J. Wang, Sens. Actuators, B, 2015, 207, 66.
18. F. K. Butt, C. Cao, W. S. Khan, M. Safdar, X. Fu, M. Tahir, F. Idrees, Z. Ali, G. Nabi and D. Yu, CrystEngComm, 2013, 15, 2106.
19. F. Nieto, A. A. Tarasenko, C. Uebing and V. Pereyra, Langmuir, 1999, 15, 5893.
20. J. Arbiol, E. Comini, G. Faglia, G. Sberveglieri and J. R. Morante, J. Cryst. Growth, 2008, 310, 253.
21. A. Andersson, J. Solid State Chem., 1982, 42, 263.
22. W. E. Slink and P. B. Degroot, J. Catal., 1981, 68, 423.
23. J. Kasperkiewicz, J. A. Kovacich and D. Lichtman, J. Electron Spectrosc. Relat. Phenom., 1983, 32, 123.
24. M. K. Takagi, M. Soma, T. Onishi and K. Tamaru, Can. J. Chem., 1980, 58, 2132.
25. H. J. Choi, J. S. Kim and M. Kang, Bull. Korean Chem. Soc., 2007, 28, 581.
26. R. Gopalakrishnan, B. V. R. Chowdari and K. L. Tan, Solid State Ionics, 1992, 53, 1168.
27. M. L. Miller and R. W. Linton, Anal. Chem., 1985, 57, 2314.
28. L. Yang, H. Lin, Z. Zhang, L. Cheng, S. Ye and M. Shao, Sens. Actuators, B, 2013, 177, 260.
29. H. H. Park, N. J. Choi, H. Kang, M. Y. Jung, J. W. park, K. H. Park and D. S. Lee, Sens. Actuators, B, 2014, 203, 282.
30. W. Jin, B. T. Dong, W. Chen, C. X. Zhao, L. Q. Mai and Y. Dai, Sens. Actuators, B, 2010, 145, 211.
31. E. K. Liu, B. S. Zhu and J. S. Luo, Semiconductor Physics, Publishing House of Electronics Industry, Beijing, 2008.
32. R. W. J. Scott, S. M. Yang, G. Chabanis, N. Coombs, D. E. Willams and G. A. Ozin, Adv. Mater., 2001, 13, 1468.
33. D. E. Williams, Sens. Actuators, B, 1999, 57, 1.
34. P. P. Sahay, J. Mater. Sci., 2005, 40, 4383.
35. G. L. Dai, Z. H. Li, J. Lu, W. N. Wang and K. N. Fan, J. Phys. Chem. C, 2012, 116, 807.
36. O. Dorosh and Z. Kisiel, Acta Phys. Pol., A, 2007, 112, S95.
37. M. Lin, T. B. Desai, F. W. Kaiser and P. D. Klugherz, Catal. Today, 2000, 61, 223.

---