Enhanced ultra-stable n-propylamine sensing behavior of V₂O₅/In₂O₃ core–shell nanorods

Abstract

Heterostructured V₂O₅/In₂O₃ core–shell nanorods were prepared by the combination of solid solution and hydrothermal methods. Microstructural and spectroscopic studies reveal the effective core–shell hetero-nanostructure. The gas sensor based on these nanorods exhibits remarkable gas sensing properties in both static and dynamic modes. It presents an optimum working temperature of 190 °C, a reasonable response speed and high selectivity with an ultra-stable reproducible response for n-propylamine. The sensor also shows enhanced optimum sensitivity (∼14), which is 3.90 times that of the pure V₂O₅ nanorods. Such promising gas sensing behavior of the hetero-nanostructured core–shell nanorods is explained using an energy band model and a suitable gas sensing mechanism has been established as well.

---

1. Introduction

Since its first conception in 1959,¹ nanotechnology has been shaping lifestyles multi-dimensionally in both basics and luxuries. Its superpositional approach can be well realized in numerous applications such as photocatalysis, batteries and gas sensors.² Solid state gas sensors have extensive uses in the detection of harmful and flammable gases, which is beneficial in domestic and industrial processing as well as in environmental monitoring.^3,4 The detection of volatile organic compounds (VOCs), such as alcohols, amines and ketones, is of crucial importance for industries, laboratories, food storage, environmental control and security purposes.⁵ However, investigations regarding the detection of VOCs below breadth limit (200 ppm) at low working temperature ranges are still rare.

Among the various types of solid state gas sensors, metal oxide semiconductor (MOS) gas sensors are of particular interest due to their cost effectiveness, portability, operation and simplicity. To date, different metal oxides (V₂O₅, In₂O₃, ZnO, and MoO₃) on the nanoscale have been investigated for the detection of various gases (ethanol, ammonia, amines, nitrogen oxides, and hydrogen sulfide).^6–12 In order to improve the gas sensing response and selectivity of the nanostructured MOS gas sensors, continuing efforts are being made, which include different methodologies, such as the formation of heterostructures by the doping and decoration of noble metals and secondary metal oxides,¹³ the formation of hybrid metal oxide nanocomposites of binary and ternary phases,^14,15 and the development of nanostructures having different morphologies.¹⁶ Among them, the casting of metal oxide heterostructures into multi-designs offers remarkable results pertaining response and selectivity. In the case of hybrid heterostructures, the enhanced gas sensing behavior is based on certain semiconducting effects, including inherent catalytic activity, provision of active surface states and lowering of depletion layers. Synthesis of core–shell hetero-nanostructures is one method, which has been performed successfully in recent years for developing a variety of metal oxides for use as gas sensors.

Earlier, several core–shell nanostructures have been successfully investigated for gas sensing, such as hydrophilic S/In₂O₃ core–shell nanocomposites,¹⁷ ZnO/In₂O₃, TeO₂/In₂O₃, and Fe₂O₃/SnO₂/Au core–shell nanocomposites.^18–20 A recent transpiring effort has been made by Li et al., who investigated Au@In₂O₃ core–shell nanospheres for sensing HCHO gas with enhanced sensitivity of 3.21 times that of In₂O₃ spheres, and the working temperature was lowered to 200 °C from 225 °C as well.²¹

Vanadium pentoxide (V₂O₅) is an important n-type transition metal oxide, which has already been successfully investigated as a gas sensing material on the nanoscale with one-dimensional morphologies.^22–26 However, its high detection threshold, elevated working temperature, low sensitivity and limited selectivity have restricted its usage in practical applications. The layered structure of V₂O₅ facilitates the intercalation of secondary metals and metal oxides on its surface or interior, which would provide additional surface reactive sites and hence improve the gas sensing behavior. Coating of external metals or metal oxides is an effective way to introduce more vacancies for oxygen adsorption and produce additional surface active states, which would result in an enhanced response. Earlier, Lu et al. synthesized polythiophene coated V₂O₅ nanotubes for sensing ethanol at room temperature.²⁷ However, the response magnitude (sensitivity) was very low (1.23), which needs further investigation.

In this study, we report the preparation of V₂O₅/In₂O₃ core–shell nanorods (V₂O₅: core, In₂O₃: shell) via a solid solution synthesis method followed by a hydrothermal reaction. In₂O₃ has been chosen, because it has already been successfully applied in core–shell heterostructures with other metal oxides, as discussed above. Coating the In₂O₃ based shell is attempted in order to increase surface active sites for increased analyte adsorption, reduction of the depletion region, and increase in the conductance, which will certainly favor their practical applications. Gas sensors based on V₂O₅/In₂O₃ core–shell nanorods are investigated to detect various VOCs (ethanol, n-propylamine, acetone, toluene and ammonia) in a wide temperature range (room temperature to 250 °C). Significant enhancement in gas sensing properties is achieved, which is explained through an energy band model of the semiconductors.

2. Experimental

2.1 Preparation of heterostructured V₂O₅/In₂O₃ core–shell nanorods

Analytical grade chemicals were used. In a typical synthesis method, precursor solution of indium was prepared in a flask by the hydrolysis of 1.25 mM indium chloride (InCl₃) in 12 mL of deionized water. Subsequently, 2.60 mM citric acid and 4 mM urea were poured into the flask while continuously stirring at 500 rpm. Strict stoichiometric ratios were maintained in this precursor solution, because they were selected as the optimal molar ratios. An acidic dispersion of 5 mM vanadium pentoxide was also prepared in 40 mL of deionized water, such that pH = 5 was maintained by the dropwise introduction of HNO₃ during its continuous stirring for 3 h.

Finally, the indium precursor solution was added slowly to the V₂O₅ dispersion during its continuous stirring for 45 min, and the resulting mixture was then transferred to an 80 mL Teflon lined stainless steel autoclave and sealed tightly. Hydrothermal reaction was carried out at 200 °C for 48 h in a conventional oven, and the mixture was then allowed to cool down to room temperature in air. Solid precipitates were separated from the suspension via centrifugation and washed subsequently with deionized water/ethanol three times and sonicated for 3 min after each time. The as-synthesized samples were dried at 60 °C and annealed at 400 °C for 2 h in air. For comparative analysis, pure V₂O₅ samples (VO) were also prepared as per the procedure described in our last work.²⁸ In order to obtain crystalline materials, the as-synthesized samples were annealed at 400 °C for 2 h and subjected to further characterization. The heterostructured V₂O₅/In₂O₃ and pure V₂O₅ samples were marked as InVO and VO, respectively.

2.2 Characterization

For structural characterization, the samples were subjected to X-ray diffraction analysis (X'Pert powder diffractometer, PANalytical, The Netherlands) using Cu K_α lines with a wavelength of 1.54 Å. Microscopic (morphological) studies of the samples were carried out by scanning electron microscope (SEM, Zeiss Ultra Plus, ZEISS, Germany) and transmission electron microscope (TEM, JEM 2100F). Oxidation states of the samples were investigated by X-ray photoelectron spectroscope (XPS, Al 300W, PE 100 eV) with an Al target. The emission angle between the photoelectron beam and the sample surface was 45°, and the calibration of the binding energy of the electron spectrometer was done by using the maximum adventitious C 1s signal at 284.6 eV with the solution of the full width at half maximum (FWHM) being 0.8 eV. Gas sensing tests were performed through a WS-30A gas sensing measurement system (Zhengzhou Winsen Technology Corp., Ltd).

2.3 Fabrication of gas sensing devices and measurement

Gas sensing devices based on both InVO and VO samples were fabricated through an in situ method. Typically, the samples were mixed with terpineol to obtain a smooth paste, which was then coated on the surface of a ceramic tube, with a pair of Au electrodes already printed on. These devices were dried at 100 °C for 10 days in air. Finally, a Ni–Cr heating wire was inserted into the ceramic tube. Before testing the gas sensing properties, the fabricated sensors were aged at 60 °C for 15 days to improve the stability of the sensitive materials. In order to compare the effects of dispersing media, a reference gas sensing device was also fabricated by replacing terpineol with water.

The gas sensing measurements were taken with a WS-30A measuring system, which comprised of a static gas chamber having a volume of 18 L. A liquid vaporization technique was used to produce the target gas molecules with a certain concentration. Five types of liquid analytes (ethanol, n-propylamine, acetone, toluene and ammonia) were taken and tested at different temperatures (room temperature (RT), 50, 80, 100, 130, 160, 190 and 250 °C).

3. Results and discussion

3.1 Structure and morphology

Comparative X-ray diffraction patterns of the InVO and VO samples are presented in Fig. 1. The diffraction pattern of InVO reveals the presence of V₂O₅ and In₂O₃ phases, and their diffraction peaks are indexed to the reference PDF patterns JCPDS no. 01-065-0131 and 01-088-2160, respectively. The most intensive peak is centered at 2θ = 20.36°, which is assigned to the (010) plane of V₂O₅ and illustrates the obvious growth along this plane. The pattern also demonstrates that the diffraction intensity of V₂O₅ is increased significantly inside InVO, which reveals its improved crystallinity. Increase in crystallinity in InVO confirms the vital role of In₂O₃ in its coverage over the V₂O₅ surface and shows reduction in grain size, which is beneficial for enhancing gas sensing performance.

Fig. 1 XRD patterns of InVO and VO nanorods.

Fig. 2(a) displays the FESEM image of the InVO sample, which clearly presents the core–shell architecture. The enlarged FESEM image of the V₂O₅-core/In₂O₃-shell nanorods (inset) shows the rod-like morphology with width ranges of 80–400 nm and lengths of up to a few micrometers. For comparison, the FESEM image of the pure V₂O₅ nanorods is shown in Fig. 2(b).

Fig. 2 (a and b) FESEM images of InVO and VO nanorods, respectively (inset: high magnification FESEM images), (c) TEM image of InVO nanorods, and (d) HRTEM image for the illustration of the surface details of InVO nanorods.

In order to have a deep insight into the structure of the InVO nanorods, TEM analysis was performed, and the results are shown in Fig. 2(c) and (d). Fig. 2(c) reveals the core–shell nanostructure, which has an estimated 30–250 nm core and 10–50 nm shell. Fig. 2(d) presents the high-resolution TEM (HRTEM) image, describing the surface details of the V₂O₅ core–In₂O₃ shell nanorods.

The spacing between two successive fringes in the core corresponds to the (010) plane of orthorhombic V₂O₅ with d ≈ 0.434 nm and in the shell corresponds to the (123) plane of In₂O₃ with d ≈ 0.273 nm.

Oxidation states of the elements were estimated through the XPS spectral studies of the sample, and the results are presented in Fig. 3. Fig. 3(a) shows the survey spectrum, indicating the presence of carbon, indium, vanadium and oxygen. Carbon (C 1s) appears due to the carbon electrodes used during the tests. Vanadium (V 2p) occurs in two states, i.e. V 2p^1/2 and V 2p^3/2, at the corresponding binding energies of 524.70 and 516.10 eV, respectively (Fig. 3(b)). Both belong to the V⁵⁺ oxidation state of V₂O₅. Indium (In 3d) occurs in two states, In 3d^3/2 and In 3d^5/2, at the corresponding binding energies of 453.45 and 445.10 eV, respectively (Fig. 3(c)). Both states are assigned to the In³⁺ state of In₂O₃. Similarly, Fig. 3(d) shows the XPS spectrum of oxygen (O 1s), which occurs at the three binding energy positions of 529.90, 531.10 and 532.4 eV, related to the oxygens in V₂O₅, In₂O₃ and adsorbed O²⁻, respectively.

Fig. 3 XPS spectrum of InVO nanorods (a) survey spectrum, (b) V 2p fitted profile, (c) In 3d fitted profile, and (d) O 1s fitted profile.

3.2 Gas sensing properties

Static mode resistance changes and the resultant response of the gas sensor based on InVO nanorods towards 200 ppm of the five target gases (ethanol, n-propylamine, acetone, toluene and ammonia) at a range of temperatures (RT to 250 °C) are presented in Fig. 4. On exposing the sensor to the targets gases, a decrease in resistance is observed, which confirms the n-type semiconducting behavior of the InVO nanorods. The results in Fig. 4(a)–(e) reveal that the sensor responds sufficiently to all five gases at 50 °C in a stable and reversible manner, which estimates it as a working temperature of the sensor based on InVO nanorods. It can also be seen that the sensor offers response that monotonically increases with increasing of the target gases' concentrations. Furthermore, at the points of target gas injection, some distortion in the signal appears in the resistance variation profiles, which is because of the manual manner by which the target gas analytes were injected into the chamber.

Fig. 4 Static mode resistance transient behavior of a gas sensor based on InVO nanorods towards 200 ppm of (a) ethanol, (b) n-propylamine, (c) acetone, (d) toluene, (e) ammonia, and (f) sensitivity response magnitude towards five test gases at different temperatures.

In order to estimate the optimum working temperatures of the sensor, the response magnitude (sensitivity) for 200 ppm of all five gases was plotted as a function of temperature, as shown in Fig. 4(f). Increase in sensitivity is noted with the rise in temperature. Maximum sensitivity is achieved at 190 °C for all the target gases except ethanol. Furthermore, the sensor exhibits the highest response towards n-propylamine among all the gases at each temperature.

Fig. 5(a) presents further illustration of the data plotted in Fig. 4(f). The highest value of sensitivity out of those for all the gases is observed for n-propylamine. In order to evaluate the enhanced sensitivity of the sensor based on InVO nanorods towards n-propylamine, a comparative analysis with respect to the gas sensor based on pure V₂O₅ (VO) nanorods was performed and is presented in Fig. 5(b). The InVO based sensor shows enhanced response compared to the VO based sensor throughout the range of temperatures. The observed enhancement in sensitivity is 1.21-fold at the working temperature (50 °C) and 3.90-fold at the optimum working temperature (190 °C). Although the optimum working temperature of the VO sensor is comparatively lower (100 °C against 190 °C), interestingly, the sensitivity at this value of temperature is also lower (Fig. 5(c)), which indeed shows the promising behavior of the InVO nanorods based sensor throughout the range of temperatures.

Fig. 5 (a) Variation of sensitivity versus temperature for InVO nanorod sensor towards all five gases, (b) comparative response analysis of InVO and VO nanorod gas sensors towards 200 ppm of n-propylamine at different temperatures, (c) estimation of the enhanced sensitivities of InVO against VO nanorod gas sensors towards 200 ppm of n-propylamine at different temperatures, and (d) evaluation of response and recovery times through real value resistance variation of gas sensor based on InVO nanorods towards 200 ppm n-propylamine at 50 °C.

In order to estimate the response speed of the InVO nanorod based sensor, its real value resistance variation was plotted for 200 ppm n-propylamine at 50 °C and is shown in Fig. 5(d). The sensor presents reasonable response speed with response and recovery times (t_resp and t_rec) of 48 s and 121 s, respectively.

Selectivity and stability (reproducibility) are the main issues for metal oxide based gas sensors. The gas sensor based on InVO nanorods shows higher selectivity towards n-propylamine compared to ethanol, acetone, toluene and ammonia, which is presented in Fig. 6. Selectivity was found by the following formula.

Fig. 6 Selectivity of InVO nanorod based gas sensor towards 200 ppm of the target gases at (a) 50 °C, and (b) 190 °C.

It was determined that at 50 °C, the sensor shows 27.61% selectivity for n-propylamine among the five target gases (Fig. 6(a)), which increases to 45.13% at the optimum working temperature of 190 °C (Fig. 6(b)). This is one of the novel characteristics of the InVO nanorod based gas sensor.

In order to probe the stability of the sensor, its repetitive response was obtained for 10 ppm of n-propylamine at 190 °C and is shown in Fig. 7. Based on the results of six repeated steps of propylamine gas injection, the reproducibility of the InVO nanorod based gas sensor was estimated by calculating the coefficient of variation (CV), which is given by the following equation.²⁶

CV = σ/μ (2)

Fig. 7 Repeated response curve of the gas sensor based on InVO nanorods for 10 ppm n-propylamine gas.

The value of CV is approximately 0.018 (1.8%), which reveals an ultra-stable and reproducible response of the InVO nanorod based sensor at its optimum working temperature.

In order to investigate the effect of the dispersing medium used during gas sensing device fabrication on the sensing performance of the InVO nanorods, comparative gas sensing trials of the gas sensors with terpineol and water were conducted, and the results are shown in Fig. 8. Fig. 8(a) presents the comparative static mode R–t curves of the sensors towards different n-propylamine concentrations at 190 °C. It is obvious that both types of sensors exhibit variation in response magnitudes at lower concentration levels, whereas no significant difference is observed at higher levels. Furthermore, the sensor with terpineol as dispersing medium shows relatively stable response compared to that of the sensor with water as the dispersing medium. This is because of the enhanced binding ability of terpineol due to its organic nature, which makes the sensing layer on the ceramic tube a bit more effective than that using water. Fig. 8(b) shows the concentration dependence of the sensitivity, which presents a similar behavior with a slight difference in magnitudes for both dispersing media.

Fig. 8 Comparative n-propylamine sensing trials of the gas sensor based on InVO nanorods at 190 °C with terpineol and water as dispersing medias (a) R–t curves, and (b) concentration dependence of corresponding sensitivities.

3.3 Gas sensing mechanism

The highly sensitive and ultra-stable gas sensing properties of the InVO nanorod based gas sensor compared with those of the pure VO based gas sensor can be explained through an energy band model. Both V₂O₅ and In₂O₃ are n-type metal oxides with distinct electronic properties, such as energy band gaps of ∼2.3 eV and ∼3 eV, work functions of ∼7 eV and ∼4.54 eV, respectively.^29,30 The formation of a heterojunction occurs at the interface of V₂O₅ (core) and In₂O₃ (shell), which is shown in the schematic energy band diagram (Fig. 9). Due to the high work function of V₂O₅, electrons flow from the In₂O₃ shell to V₂O₅ core until the equilibrium of Fermi levels is established, which results in formation of an electron depletion layer at the interface. A band bending occurs, which leads to a heterojunction barrier. The exposure of the sensor to target gases may produce variation in this heterojunction barrier, which is responsible for the enhanced gas sensing performance. This type of effect has also been studied by others.³¹

When the sensor is exposed to air, trapped electrons in the In₂O₃ shell result in increasing of the sensor's resistance (base-line value). However, on exposure to target gases, these depleted electrons are released back to the conduction band of the In₂O₃ shell, which contributes to reduce the heterojunction barrier at the contact of In₂O₃–V₂O₅ and leads to a drop in resistance value. Such a variation in resistance is supported from the response curves in both the static and dynamic modes.

Enhancement in the gas sensing characteristics of the InVO nanorod based gas sensor is also due to a synergistic effect. We know that the Debye length is <30 nm for metal oxides.³² For a synergistic effect in improved gas sensing parameters, the shell's thickness must be less than or close to the Debye length.³³ Since the shell thickness in the InVO core–shell nanorods used in this study lies in the range of 10–50 nm, which comparable is to the Debye length, the enhanced gas sensing performance is accounted for on the basis of this valid hypothesis. A schematic of the sensing mechanism is displayed in Fig. 9.

Fig. 9 Schematic of the gas sensing mechanism of the gas sensor based on InVO nanorods in (a) air and (b) exposed to target gas.

The higher selectivity of the InVO nanorod sensor towards n-propylamine is based on the selective oxidation of n-propylamine and can be explained on the basis of selective amino-oxidation mechanism of the hetero-catalytic nature of the nanorods. The intermediate compound obtained during the hydrothermal reaction exhibits vanadic acid-like behavior because of the functionalization of the precursor solution with H₂O₂ and HNO₃ acids. It is known that the selective oxidation of primary amines is favorable in the presence of vanadic acids.³⁴ In addition, the promotion of adsorption and oxidation of n-propylamine gas due to the presence of the two types of centers with different reduction–oxidation and acid–base properties are also conceivable causes of higher selectivity.³⁵

4. Conclusions

A facile combination of solid solution and hydrothermal methods has been applied to the preparation of heterostructured V₂O₅/In₂O₃ core–shell (InVO) nanorods. The gas sensor for VOCs (ethanol, n-propylamine, acetone, toluene and ammonia) based on the InVO nanorods exhibits low and optimum working temperatures of 50 and 190 °C, respectively. The sensor also presents high selectivity for n-propylamine gas at all of temperatures with enhanced reproducible optimum sensitivity (∼14), which is 3.90 times higher than that of the pure V₂O₅ nanorods. The InVO nanorod based sensor shows reasonable response (48 s) and recovery times (121 s) at 50 °C for 200 ppm n-propylamine. Furthermore, it shows a remarkable selectivity (53.81% for 200 ppm) with ultra-stable reproducibility (coefficient of variation: 1.8% for 10 ppm) for n-propylamine at its optimum working temperature, which ascertains its significance for industrial and environmental applications.

Acknowledgements

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

References

1. R. P. Feynman, J. Microelectromech. Syst., 1992, 1, 60.
2. X. Z. Feng, A. M. Liu and J. W. Cheng, Adv. Mater. Res., 2010, 121–122, 5.
3. G. Sberveglieri, Sens. Actuators, B, 1995, 23, 103.
4. A. M. Azad, S. A. Akbar, S. G. Mhaisalkar, L. D. Birkefeld and K. S. Goto, J. Electrochem. Soc., 1992, 139, 3690.
5. Z. Li, Q. Zhao, W. Fan and J. Zhan, Nanoscale, 2011, 3, 1646.
6. V. Modafferi, G. Panzera, A. Donato, P. L. Antonucci, C. Cannilla, N. Donato, D. Spadaro and G. Neri, Sens. Actuators, B, 2012, 163, 61.
7. W. Jin, W. Chen, B. T. Dong, C. X. Zhao, L. Q. Mai and Y. Dai, Key Eng. Mater., 2009, 421–422, 328.
8. L. Gao, Z. Cheng, Q. Xiang, Y. Zhang and J. Xu, Sens. Actuators, B, 2015, 208, 436.
9. H. Ibrahim and Y. Temerk, Sens. Actuators, B, 2015, 206, 744.
10. P. S. Venkatesh, P. Dharmaraj, V. Purushothaman, V. Ramakrishnan and K. Jeganathan, Sens. Actuators, B, 2015, 212, 10.
11. M. Yin and S. Liu, Sens. Actuators, B, 2015, 209, 343.
12. C. Imawan, F. Solzbacher, H. Steffes and E. Obermeier, Sens. Actuators, B, 2000, 64, 193.
13. W. Jin, B. Dong, W. Chen, C. Zhao, L. Mai and Y. Dai, Sens. Actuators, B, 2010, 145, 211.
14. K. Galatsis, Y. X. Li, W. Wlodarski, E. Comini, G. Sberveglieri, C. Cantalini, S. Santucci and M. Passacantando, Sens. Actuators, B, 2002, 83, 276.
15. Y. V. Kaneti, J. Moriceau, M. Liu, Y. Yuan, Q. Zakaria, X. Jiang and A. Yu, Sens. Actuators, B, 2015, 209, 889.
16. Y. Xia, P. Yang, Y. Sun, Y. Wu, B. Mayers, B. Gates, Y. Yin, F. Kim and H. Yan, Adv. Mater., 2003, 15, 353.
17. S. Meng, Z. Cao, X. Fu and S. Chen, Appl. Surf. Sci., 2015, 324, 188.
18. Y. L. Zhao, H. X. Lu, X. J. Yu, B. B. Fan, D. L. Chen, L. W. Zhang, H. L. Wang, D. Y. Yang, H. L. Xu and R. Zhang, Appl. Surf. Sci., 2011, 257, 10634.
19. H. Ko, S. Park, S. An and C. Lee, Curr. Appl. Phys., 2013, 13, 919.
20. X. Liu, J. Zhang, X. Guo, S. Wang and S. Wu, RSC Adv., 2012, 2, 1650.
21. X. Li, J. Liu, H. Guo, X. Zhou, C. Wang, P. Sun, X. Hu and G. Lu, RSC Adv., 2015, 5, 545.
22. I. Raible, M. Burghard, U. Schlecht, A. Yasuda and T. Vossmeyer, Sens. Actuators, B, 2005, 106, 730.
23. A. D. Raj, T. Pazhanivel, P. S. Kumar, D. Mangalaraj, D. Nataraj and N. Ponpandian, Curr. Appl. Phys., 2010, 10, 531.
24. T. Zhai, H. Liu, H. Li, X. Fang, M. Liao, L. Li, H. Zhou, Y. Koide, Y. Bando and D. Golberg, Adv. Mater., 2010, 22, 2547.
25. J. Liu, X. Wang, Q. Peng and Y. Li, Adv. Mater., 2005, 17, 764.
26. S. A. Hakim, Y. Liu, G. S. Zakharova and W. Chen, RSC Adv., 2015, 5, 23489.
27. Y. Lu, W. Jin and W. Chen, Mater. Sci. Forum, 2010, 654–656, 1154.
28. S. A. Hakim, Y. Liu, Y. Lu and W. Chen, Mater. Sci. Semicond. Process., 2015, 31, 630.
29. A. Gurlo, N. Bârsan, M. Ivanovskaya, U. Weimar and W. Göpel, Sens. Actuators, B, 1998, 47, 92.
30. D. W. Bullett, J. Phys. C: Solid State Phys., 1980, 13, L595.
31. T. Tharsika, A. Haseeb, S. Akbar, M. Sabri and W. Hoong, Sensors, 2014, 14, 14586.
32. S. Park, S. An, Y. Mun and C. Lee, ACS Appl. Mater. Interfaces, 2013, 5, 4285.
33. Y. J. Chen, G. Xiao, T. S. Wang, F. Zhang, Y. Ma, P. Gao, C. L. Zhu, E. Zhang, Z. Xu and Q. H. Li, Sens. Actuators, B, 2011, 156, 867.
34. J. S. Reddy and A. Sayari, Appl. Catal., A, 1995, 128, 231.
35. K. W. Kim, P. S. Cho, S. J. Kim, J. H. Lee, C. Y. Kang, J. S. Kim and S. J. Yoon, Sens. Actuators, B, 2007, 123, 318.

---

Footnote

† These authors contributed equally to this study and share first authorship.

---