Three dimensions sphere formaldehyde nanosensor applications: preparation and sensing properties

Abstract

Zn@SnO₂ microspheres with hierarchical structure were prepared through a simple solvothermal method; the structure and morphology were characterized by X-ray diffraction (XRD), scanning electron microscopy (SEM), and high resolution transmission electron microscopy (HRTEM) showing the materials with extraordinary 3D nanoarchitectures. The gas sensing properties of the as-prepared pure SnO₂ and Zn-doped SnO₂ were tested toward various gases. The results showed that the SnO₂ sensor with 6.67 wt% Zn-doping displayed an excellent selectivity toward formaldehyde at the operating temperature 160 °C, which was considerably lower than most formaldehyde sensors in heater type among previous reports, in addition to giving a response of about 15.2 to 100 ppm, which is about 2.1 times higher than that of sensors based on pure SnO₂. The τ_res and the τ_rec values of the 6.67 wt% Zn-doped SnO₂ sensor to 100 ppm formaldehyde were 2 s and 2 s respectively, demonstrating extraordinary gas sensing properties, whereas those of the pure SnO₂ sensor were relatively long. The enhancement might be attributed to the unique morphology and increased oxygen vacancy due to the Zn doping.

---

Introduction

As we all know, fitment pollution is one of the most serious and widespread air pollution sources indoors, while many necessities such as upholstery, carpeting, and plastic products give out volatile organic compounds (VOCs), including formaldehyde, which is paramount as a cause of sick building syndrome SBS.^1,2 The World Health Organization (WHO) and Chinese Environmental Protection Administration have established standards to limit exposure to VOCs, particularly formaldehyde vapor. In the past decades, numerous analytical methods for the detection of VOCs have been reported, using spectrophotometry,³ gas chromatography (GC),⁴ high-performance liquid chromatography⁵ and ion chromatography,⁶ which are usually costly with bulky instruments requiring high power and complex operation procedures and are unable to provide VOCs exposure information on a real-time basis. Therefore, the detection of gaseous formaldehyde (HCHO) is so difficult, and thus it is still highly challenging to quantify and monitor.

To maintain environmental safety, gas sensors are required for examining poisonous and hazardous gases such as ethanol, acetone, toluene and HCHO, which have been widely applied in the fields of healthcare, safety, environmental monitoring and chemical process control since the introduction of chemical sensors in 1922.^7–11 The past few decades have witnessed the development of various sensors, which are closely related to both species and morphologies of the gas sensing materials. With the rapid and continuous development of nanoscience and technology, considerable efforts have been made to synthesize metal oxide semiconductors (MOSs), such as In₂O₃,¹² Co₃O₄,¹³ Cr₂O₃,¹⁴ NiO,¹⁵ ZnO,¹⁶ and SnO₂,¹⁷ with different morphologies such as nanofiber, core–shell, flower-like and hierarchical nanostructures by the convenience of a rich variety of synthesis routes available (such as electrospinning, hydrothermal synthesis, solvothermal routes and vapor–solid). Among them, SnO₂, with a direct band gap of 3.6 eV, has been extensively used as gas sensing materials to detect gases, such as ethanol, CO, NO₂, and H₂S, due to its low resistance, strong interaction with gas molecules and well controlled morphology.

In this study, Zn@SnO₂ microspheres formed by 2D nanosheets with hierarchical structure were successfully fabricated by introducing Zn into n-type SnO₂ via one-step solvothermal method. Their super formaldehyde sensing properties were investigated via a convenient and lossless measurement technique, which was reported for the first time. Highly efficient sensing performance against formaldehyde was observed, which is considerably better than published reports, making the fabricated nanomaterial a good candidate sensing material for high performance formaldehyde sensors.

Experimental section

Materials synthesis

All the used chemical reagents in this work were of analytical grade and used as purchased without further purification: SnCl₂·2H₂O, Zn(NO₃)₂·6H₂O, NaOH, and Na₃C₆H₅O₇·2H₂O.

In a typical procedure, Zn@SnO₂ microspheres were synthesized by a typical solvothermal method. First, a certain volume of ethanol (15 mL) was dissolved in deionized water (15 mL) to form a clear solution, and then SnCl₂·2H₂O (0.9 g), NaOH (4 mmol), sodium citrate (1 mmol) and different amounts of Zn (NO₃)₂·6H₂O were added into the abovementioned solution under constant stirring. After being stirred vigorously for 5 h, the solutions were transferred into 50 mL Teon-lined stainless steel autoclaves. The autoclaves were heated at 180 °C for 10 h and then cooled to room temperature naturally. The precipitates were separated by centrifugation, washed with distilled water and ethanol five times, and dried at 60 °C for 24 h. Finally, microspheres with hierarchical structure were obtained. We refer to these samples as S0, S1, S2, and S3, representing pure SnO₂, and the Zn/SnO₂ weight percentage of 0, 6.67, 3.33 and 10, respectively.

Characterization

X-Ray diffraction (XRD) analysis was conducted using a Scintag XDS-2000 X-ray diffractometer with Cu Kα radiation (λ = 1.5418 Å) to analyze the structure of the prepared products. Scanning electron microscopy (SEM) images were performed using a SHIMADZU SSX-550 (Japan) instrument to observe the morphology of the prepared products. Transmission electron microscope (TEM) images were obtained using a JEM-ARM200F microscope to observe detailed microstructures and detect elements of the prepared products.

Fabrication and measurement of gas sensor

The as-prepared material was mixed with deionized water in a weight ratio of 100 : 25 and ground in a mortar for 3 h to form a paste. The paste was then coated on an Al₂O₃ ceramic tube to form a sensing film (a thickness of about 300 μm) on which a couple of parallel Au electrodes were previously printed. Pt lead wires attached to these Au electrodes were used as electrical contacts. After the ceramic tube was calcined at 300 °C for 2 h, a Ni–Cr heating wire was inserted into the ceramic tube as a heater for controlling the operating temperature. The structure of a sensor is shown in Fig. 1. The details of the sensor fabrication were similar to our previous works.¹³

Fig. 1 (a) Illustration of a ceramic tube coated with sensing material; (b) schematic structure of a completed gas sensor.

Gas sensing properties were measured by a CGS-8 intelligent gas sensing analysis system (Beijing Elite Tech Co., Ltd., China) under laboratory conditions (25 °C, 40 RH%). The test gases were injected into a test chamber with a microinjector. The response value (S) was defined as S = R_a/R_g, where R_a and R_g denoted the sensor's resistance in the air and presence of the target gases, respectively. The time taken by the sensor to achieve 90% of the total resistance change was defined as response time when the target gas was introduced to the sensor (target gas adsorption) or the recovery time when the chamber was full of air replacing target gas (target gas desorption).¹⁸

Results and discussion

Structural and morphological characteristics

The XRD patterns of Zn-doped SnO₂ (S1) are shown in Fig. 2. It can be observed that all of the diffraction peaks of the products can be indexed to SnO₂, and no other peaks corresponding to impurities were observed, which was consistent with the Joint Committee on Powder Diffraction Standards card (JCPDS, 41-1445). The peak of ZnO or any other compound including Zn is not detected in Fig. 2, manifesting the place in the lattice atom of Sn⁴⁺ that may be substituted by Zn²⁺ from another perspective.

Fig. 2 XRD patterns of Zn@SnO₂ (S1).

To get further information about the unique nanoarchitecture, SEM and HRTEM analyses were performed. The morphologies and nanostructures of the as-prepared Zn-doped SnO₂ (S1) products were characterized using FESEM and HRTEM, as shown in Fig. 3(a) and (b); displayed low magnification SEM images of the Zn@SnO₂ (S1); (c) and (d) displayed the high magnification SEM images of the Zn-doped SnO₂ (S1) in different directions; (e–g) displayed HRTEM images of the as-prepared Zn@SnO₂ (S1) and two independent enlarged crystalline structure of the nanosheet. No other morphologies could be detected in Fig. 3(a), and it is clear from the image that all the 3D structures were sphere-like architectures, with a uniform diameter of about 7 μm. Fig. 3(b) is the enlarged FESEM of a single sphere, showing that the sphere-like architectures consist of many nanosheets, and estimated a diametral quotient in 7 μm, which further confirm the result obtained from Fig. 3(a). The high-magnification FESEM (Fig. 3(c) and (d)) shows the detailed morphological information of sample S1, and it reveals that the diameter and thickness of the sphere nanostructure are about 7.56 μm and 80 nm, implying that the overall size of each sphere ranges from 7 ± 0.5 μm horizontally and 80 ± 5 nm vertically, which is well consistent with estimations of (Fig. 3(a) and (b)). The crystalline nature of these Zn@SnO₂ are confirmed from transmission electron microscopy (TEM) analysis, which shows clear lattice fringes, as shown in Fig. 3(f–g). Fig. 3(e) is one of the as-prepared hierarchical architectures self-organized nanosheet. Area (A) surrounded by the blue dotted line corresponded to Fig. 3(f), and the red dotted line encapsulated area (B) referred to Fig. 3(g), respectively. The lattice fringes are at a distance of 2.37 Å, as shown in Fig. 3(f), which is in good agreement with the d spacing value of the (200) plane of SnO₂, while the lattice spacing of 3.47 Å, as shown in Fig. 3(g), corresponded to the (110) plane of the SnO₂ crystal, revealing that these SnO₂ nanosheets preferentially grew along [200] (Fig. 3f) and [110] direction, (Fig. 3(g)), which is in accordance with XRD results of rutile Fig. 2, confirming the formation of crystalline structure of Zn@SnO₂.

Fig. 3 (a and b) Low magnification SEM images of the Zn@SnO₂ (S1); (c and d) low magnification SEM images of the Zn@SnO₂ (S1); and (e–g) TEM images of the Zn@SnO₂ (S1).

Gas sensor performance and sensing mechanism

We all know that the gas response of a semiconductor sensor is usually dependent on the sensor's operating temperature.^13,27 The responses of the sensors based on S0, S1, S2 and S3 to 100 ppm formaldehyde (HCHO) were tested to determine the optimum operating temperature, as shown in Fig. 4. It can be observed that the responses of the tested sensor varied with operating temperature. According to Fig. 4, 160 °C was suggested to be the optimum operating temperature for formaldehyde detection based on S0 (pure SnO₂) and S1 (6.67 wt% Zn-doped SnO₂) sensors, because these sensors showed the maximum response of 7.2 and 15.2 at the corresponding temperature. While 205 °C was identified as the optimum operating temperature for formaldehyde detection based on S2 (3.33 wt% Zn-doped SnO₂) and S3 (10 wt% Zn-doped SnO₂) sensors, because the two sensors increased rapidly and showed the maximum response of 7.64 and 5.39 at the corresponding temperature followed by a decrease with increasing operating temperature. It is evident that the S1 sensor showed the highest response to formaldehyde at a lower corresponding optimum operating temperature of 160 °C compared to S2, S3. Therefore, we will just talk about the S0 and S1 sensors in the following discussion. Furthermore, our gas sensor has a relatively low working temperature compared with those formaldehyde sensors based on Fe₂O₃–In₂O₃ nanotubes (250 °C),¹⁹ CuO nanocubes (300 °C),²⁰ and ZnO nanocones (275 °C).²¹ Operating temperature comparison of various sensors toward formaldehyde is shown in Table 1.

Fig. 4 Responses of sensors based on S0, S1, S2 and S3 to 100 ppm formaldehyde as a function of operating temperature.

Table 1 Operating temperature comparison of various gas sensors toward formaldehyde

Material	Preparation method	Concentration (ppm)	Operating temperature (°C)
SnO2/In2O3 (ref. 22)	Electrospinning	10	375
Co3O4 (ref. 23)	Hydrolysis and condensation	3	220
In2O3 (ref. 24)	Chemical spray pyrolysis	80	275
LaFeO3 (ref. 25)	PMMA crystal template	100	190
SnO2 (ref. 26)	Topological transformation	100	330
Fe3O4@Co3O4 (ref. 27)	Hydrothermal	100	240
Zn@SnO2 in this paper	Solvothermal	100	160

---

The reason why the sensor based on S1 shows the highest response to formaldehyde at 160 °C may be as follows: the increase in operating temperature could facilitate the chemical reaction, which will lead to the increase of response. Moreover, the constantly rising temperature can potentially accelerate desorption of the HCHO molecules, while hindering adsorption of the HCHO molecules. When adsorption and desorption of HCHO molecules attained dynamic equilibrium, it reached the best working temperature. Thus, the change in resistance would be decreased when working temperature reached a higher level. Namely, the sensitivity would be reduced. Optimum temperature illustration of the sensor based on S1 is shown in Fig. 5.

Fig. 5 Optimum temperature illustration of sensors based on S1.

Response and recovery times are important parameters for gas sensors. Fig. 6 shows the dynamic response of the sensor based on S0 (pure SnO₂) and S1 (6.67 wt% Zn-doped SnO₂) to 100 ppm formaldehyde at their own optimum working temperatures. It can be seen that the response and recovery rate of the S1 (6.67 wt% Zn-doped SnO₂) sensor is considerably higher than that of the S0 (pure SnO₂). The τ_res value of the S1 sensor is very short (2 s to 100 ppm formaldehyde), while that of the pure SnO₂ sensor was relatively long. The τ_rec value of the S1 sensor is also shorter (2 s to 100 ppm formaldehyde) than that of the pure SnO₂ sensor, which is far more rapid than most reported formaldehyde sensors, such as the TiO₂ nanocube sensor²⁸ (response time 180 s, recovery time 780 s), Au@SnO₂ sensor¹⁷ (response time 80 s, recovery time 62 s). Response and recovery time comparisons of various formaldehyde sensors are displayed in Table 2.

Fig. 6 Responses to 100 ppm formaldehyde for the sensor based on S0 and S1 at 160 °C.

Table 2 Comparisons of response and recovery time of various gas sensors toward formaldehyde

Material	Preparation method	Concentration (ppm)	Response time (s)	Recovery time (s)
SnO2/In2O3 (ref. 22)	Electrospinning	10	20	40
SnO2 (ref. 29)	Acid-washing	100	13	35
LaFeO3 (ref. 25)	PMMA crystal template	100	25	23
La0.7Sr0.3FeO3 (ref. 30)	Electrospinning	20	220	100
ZnO^31	Thermostat water bath	100	6.5	24
Au@SnO2 (ref. 17)	Sol–gel	50	80	62
NiO^32	Hydrothermal	100	50	150
LaFeO3 (ref. 33)	Hydrothermal	500	13	25
Zn@SnO2 in this paper	Solvothermal	100	2	2

---

Moreover, when the target gas was injected into the testing chamber, the responses of both sensors increased rapidly; when subjected to air, the sensor recovered to its initial state rapidly. The quick response and recovery of the sensor may be attributed to the hierarchical nanostructure and special morphologies.

To demonstrate the performance of the S1 (6.67 wt% Zn-doped SnO₂) sensor as an excellent sensing material, the relationship between the response and formaldehyde concentrations at operating temperature of 160 °C is depicted in Fig. 7. It is seen from the curve that the response increased with the increasing of formaldehyde concentration from 1 to 100 ppm. As the concentration of formaldehyde rose, the responses increased. The response was about 1.36, 2.5, 7.9, 10.5, 12.8 and 15.2 to 1, 5, 10, 20, 50 and 100 ppm formaldehyde.

Fig. 7 Responses of sensors based on Zn@SnO₂ at 160 °C versus formaldehyde concentrations.

The response of sensors using pristine SnO₂ and hierarchical 6.67 wt% Zn-doped SnO₂ samples versus the formaldehyde concentration ranging from 1 to 2000 ppm at 160 °C is shown in Fig. 8. From the curve, it is found that the responses of both sensors increased rapidly with increase in the formaldehyde concentration (1–100 ppm, Fig. 8(a)) and then gradually tended to saturation when the formaldehyde concentrations reached higher levels (100–2000 ppm Fig. 8(b)). Evidently, the sensor based on S1 exhibited the higher response to formaldehyde at different concentrations compared with that based on S0 in the whole detecting range. Fig. 8 clearly shows that introduction of Zn in SnO₂ improved its sensing performance in terms of response, indicating that S1 (6.67 wt% Zn-doped SnO₂) possesses a more sensitive property to formaldehyde.

Fig. 8 (a) The responses of sensors based on pure SnO₂ and Zn@SnO₂ versus formaldehyde concentrations in the range of 1–100 ppm. (b) The responses of sensors based on pure SnO₂ and Zn@SnO₂ versus formaldehyde concentrations in the range of 1–2000 ppm.

The gas sensing selectivity is another important parameter to evaluate the sensing ability of semiconductor materials. Fig. 9 shows the cross responses of sensors based on pristine SnO₂ and hierarchical 6.67 wt% Zn-doped SnO₂ samples to a variety of gases with the concentration of 100 ppm, which were tested at their optimum operating temperatures of 160 °C. It can be observed that the response of hierarchical 6.67 wt% Zn-doped SnO₂ to 100 ppm HCHO at 160 °C was 15.2, which was higher than the response to 100 ppm CO, NH₃, NO₂, C₇H₈, C₈H₁₀, CH₃OH, C₃H₆O (1.02–2.68) (Table 3). For the pristine SnO₂ sensor, the R_a/R_g value to 100 ppm HCHO at 160 °C increased to 7.2, while those to most of the other gases decreased (1.03–4.46) (Table 3). The selectivity to target gases was defined as the ratio of gas response to 100 ppm target gases and that to other gases S_HCHO/S_gas. S_HCHO/S_gas values of interference gases were 1.61–6.99 for the pristine SnO₂ sensor (Table 3). These values increased to 5.67–14.9 for hierarchical 6.67 wt% Zn-doped SnO₂ sensor. Therefore, these results clearly demonstrated that the sensor based on S1 (6.67 wt% Zn-doped SnO₂) was effective for selectivity toward formaldehyde.

Fig. 9 Responses of sensors based on pure SnO₂ and Zn@SnO₂ at 160 °C to 100 ppm various gases.

Table 3 Response and S_HCHO/S_gas comparisons of S0 and S1 gas sensors toward various gases

Gas	Ra/Rg S0	SHCHO/Sgas S0	Ra/Rg S1	SHCHO/Sgas S1
CO	1.03	6.99	1.02	14.9
NH3	1.13	6.37	1.04	14.6
NO2	1.33	5.41	1.18	12.9
C7H8	1.46	4.93	1.42	10.7
C8H10	3.03	2.38	1.87	8.12
CH3OH	4.28	1.68	2.28	6.67
C3H6O	4.46	1.61	2.68	5.67
HCHO	7.2	1	15.2	1

---

The better sensitivity for formaldehyde than other gases is considered to be caused by the different optimum working temperatures of the sensor to different gases. According to previous reports, the sensor showed selectivity at different operating temperatures due to the distinction of the orbital energy of the gas molecule. Therefore, the nanomaterial showed higher sensitivity to HCHO over other gases.

Gas sensing mechanism

Typically, the sensing mechanism could be explained through the change in the resistance of the sensor caused by the adsorption and desorption process of gas molecules on the surface of the oxide.^34–36 When 6.67 wt% Zn-doped SnO₂ sensor is surrounded by air, oxygen molecules adsorb onto its surface to generate chemisorbed oxygen species (O₂⁻, O⁻, O²⁻),³⁷ which can lead to a decrease in conductivity and an increase in resistance of the sensor. The oxygen molecules capture electrons from Zn-doped SnO₂ and transforms into O⁻ (420–670 K), O²⁻ (above 670 K), and O₂⁻ (below 420 K) at the surface of the sensing layer, which would lead the holes to spread all over the surface of Zn-doped SnO₂, and the resistance increased.³⁸ When the sensor is exposed to reducing HCHO gas, HCHO molecules can react with the chemisorbed oxygen species and the trapped electrons are released back to the conduction band of SnO₂, which will increase the carrier concentration and electron mobility, resulting in the reduction of sensor resistance (schematic illustration is shown in Fig. 10). The processes of the reaction can be described as follows:

O₂ (gas) + 2e⁻ (ads) → 2O⁻ (ads) (1)

HCHO (gas) + 2O⁻ (ads) → CO₂ + H₂O + 2e⁻ (2)

Fig. 10 Schematic of gases' conductive processes in the sensor system: (1) in air; (2) oxygen molecules trap electrons; (3) chemisorbed oxygen; (4) HCHO reacts with absorbed surface chemisorbed oxygen.

Formaldehyde molecules react with absorbed surface oxygen and release electrons back to the conduction band of SnO₂, thus increasing the carrier concentration and carrier mobility of Zn-doped SnO₂ and leading to a decrease in the sensor resistance.

The quick response should be understood in the framework of gas diffusion toward the SnO₂ surface and its reaction with surface oxygen species. At a stationary temperature, the surface reaction between HCHO and adsorbed oxygen will not vary significantly. Accordingly, gas diffusion could be regarded as the key factor. When the primary nanosheets are agglomerated to a significant extent, the target gas begins to interact with the nanosheets located at the outer part of the SnO₂ sphere. In addition, the diffusion of the target gas toward the sensing surface of the interior particles requires a homologous time. First, on the one hand, the broad surfaces of nanosheets in the unique 3D spherical nanostructure become more reactive and likely to absorb oxygen and form ionized oxygen species, which can facilitate fast mass transfer of HCHO molecules to and from the interaction area as well as improve the rate for electrons. On the other hand, the amount of oxygen that can be absorbed and ionized is increased, which is induced by the increased specific area of the Zn-doped SnO₂. Thus, extraordinary architecture could be considered as a factor that promotes the device to absorb more HCHO molecules on the sensor. Moreover, the broad surface of nanosheets makes the absorption of oxygen and the HCHO molecule easier, resulting in super-speed response and recovery.

Second, when the Zn-doped SnO₂ 3D spherical nanostructure was introduced to the reductive HCHO atmosphere, the substitution of Zn²⁺ for Sn⁴⁺ could enhance the active surface to more generation of surface oxygen vacancies according to the solid state chemistry theory, resulting in higher sensor responses. Furthermore, the zinc replacement of tin in the SnO₂ lattice may form certain amounts of catalytically active centers,³⁹ which work effectively as centers of the oxygen chemisorption and reducing gases oxidation and provide electronic exchange between absorbed species and the SnO₂ conduction band, leading to the optimization of adsorption/desorption dynamic equilibrium at the SnO₂ surface, controlling gas sensing effects, and contributing to changes of gas sensing properties (a schematic presentation of the explanation of gas sensing enhancement is displayed in Fig. 11). Certainly, our group is still in further research of the mechanism.

Fig. 11 Diagram illustrating the influence of zinc substitution of tin lattice dispersed SnO₂ phase on the area of intergrain contacts.

Conclusions

In summary, pure SnO₂ and doped SnO₂ with different Zn content were synthesized through a solvothermal method, and their HCHO sensing properties were investigated. The results showed that Zn-doping can improve their HCHO sensing performance greatly compared to pure SnO₂. The SnO₂ sensor with 6.67 wt% Zn doping exhibited the highest response and best selectivity among all the samples. The improvement of sensing properties was attributed to the increase of oxygen vacancy induced by the Zn doped and extensive surface area of the 3D SnO₂ sphere, which has been well activated and utilized for gas sensing.

Acknowledgements

The authors are grateful to the National Natural Science Foundation of China (Grant 61274068, 61370046), the National High Technology Research and Development Program of China (Grant no. 2013AA030902), Project of Science and Technology Development Plan of Jilin Province (Grant nos 20130206021GX, 20140204056GX), and Project of Science and Technology Plan of Changchun City (Grant no. 14KG019, 13KG49, 14KG020).

Notes and references

1. Po-R. Chung, C.-T. Tzeng, M.-T. Ke and C.-Y. Lee, Sensors, 2013, 13, 4468–4484.
2. W. J. Kim, N. Terada, T. Nomura, R. Takahashi, S. D. Lee, J. H. Park and A. Konno, Clin. Exp. Allergy, 2002, 32, 287–295.
3. G. R. Mohimann, Appl. Spectrosc., 1985, 39, 98–101.
4. T. Dumas, J. Chromatogr., 1982, 247, 289–295.
5. B. Mann and M. L. Grayeski, J. Chromatogr. A, 1987, 386, 149–158.
6. J. M. Lorrain, C. R. Fortune and B. Dellinger, Anal. Chem., 1981, 53, 1302–1305.
7. M. Mascini and C. Cremisini, Anal. Chim. Acta, 1977, 92, 277–328.
8. S. Maheshwari and H.-C. Chang, Adv. Mater., 2009, 21, 349–354.
9. X. Peng, Acc. Chem. Res., 2010, 43, 1387–1395.
10. M. D. Regulacio and M. Y. Han, Acc. Chem. Res., 2010, 43, 621–630.
11. D. D. D. Ma, C. S. Lee, F. C. K. Au, S. Y. Tong and S. T. Lee, Science, 2003, 299, 1874–1877.
12. J. Liu, W. Guo, F. Qu, C. Feng, C. Li, L. Zhu, J. Zhou, S. Ruan and W. Chen, Ceram. Int., 2014, 40, 6685–6689.
13. F. Qu, J. Liu, Y. Wang, S. Wen, Y. Chen, X. Li and S. Ruan, Sens. Actuators, B, 2014, 199, 346–353.
14. H. Liu, X. Du, X. Xing, G. Wang and S. Z. Qiao, Chem. Commun., 2012, 48, 865–867.
15. P. Wu, J. H. Sun, Y. Y. Huang, G. F. Gu and D. G. Tong, Mater. Lett., 2012, 82, 191–194.
16. P. Feng, Q. Wan and T. H. Wang, Appl. Phys. Lett., 2005, 87, 213111.
17. F.-C. Chung, R.-J. Wu and Fu-C. Cheng, Sens. Actuators, B, 2014, 190, 1–7.
18. C. Feng, W. Li, C. Li, L. Zhu, H. Zhang, Y. Zhang, S. Ruan, W. Chen and L. Yu, Sens. Actuators, B, 2012, 166–167, 83–88.
19. X. Chi, C. Li, L. Liu, S. Li, H. Li, X. Zhang, X. Bo and H. Shan, Mater. Sci. Semicond. Process., 2014, 18, 160–164.
20. H. Ju Park, N.-J. Choi, H. Kang, M. Y. Jung, J. W. Park, K. H. Park and D.-S. Lee, Sens. Actuators, B, 2014, 203, 282–288.
21. S. Hussain, T. Liu, M. Kashif, S. Cao, W. Zeng, S. Xu, K. Naseer and U. Hashim, Mater. Lett., 2014, 128, 35–38.
22. H. Du, J. Wang, Y. Sun, P. Yao, X. Li and N. Yu, Sens. Actuators, B, 2015, 206, 753–763.
23. J. Y. Kim, N.-J. Choi, H. J. Park, J. Kim, D.-S. Lee and H. Song, J. Phys. Chem. C, 2014, 118, 25994–26002.
24. N. G. Pramod and S. N. Pandey, Ceram. Int., 2015, 41, 527–532.
25. J. Qin, Z. Cui, X. Yang, S. Zhu, Z. Li and Y. Liang, Sens. Actuators, B, 2015, 209, 706–713.
26. K. Xu, D. Zeng, J. Wu, Q. Mao, S. Tian, S. Zhang and C. Xie, Sens. Actuators, B, 2014, 205, 416–425.
27. F. Qu, J. Liu, Y. Wang, S. Wen, Y. Chen, X. Li and S. Ruan, Sens. Actuators, B, 2014, 199, 346–353.
28. G. Wu, J. Zhang, X. Wang, J. Liao, H. Xia, S. A. Akbar, J. Li, S. Lin, X. Li and J. Wang, Ceram. Int., 2012, 38, 6341–6347.
29. J. Huang, L. Wang, C. Gu, Z. Wang, Y. Sun and J.-J. Shim, Sens. Actuators, B, 2015, 207, 782–790.
30. P.-J. Yao, J. Wang, Q. Qiao and H.-ying Du, J. Mater. Sci., 2015, 50, 1338–1349.
31. J. Zhao, C. Xie, L. Yang, S. Zhang, G. Zhang and Z. Cai, Appl. Surf. Sci., 2015, 330, 126–133.
32. X. Sun, X. Hu, Y. Wang, R. Xiong, X. Li, J. Liu, H. Ji, X. Li, S. Cai and C. Zheng, J. Phys. Chem. C, 2015, 119, 3228–3237.
33. H. Zhang, P. Song, D. Han and Q. Wang, Physica E, 2014, 63, 21–26.
34. J. Park, X. Shen and G. Wang, Sens. Actuators, B, 2009, 136, 494–498.
35. J. Xu, X. Wang, G. Wang, J. Han and Yu’an Sua, Electrochem. Solid-State Lett., 2006, 9(11), H103–H107.
36. B. Varghese, T. Choon Hoong, Z. Yanwu, M. V. Reddy, V. Bobba, R. Chowdari, A. T. S. Wee, T. B. C. Vincent, C. Teck Lim and C.-H. Sow, Adv. Funct. Mater., 2007, 17, 1932–1939.
37. X. Xu, H. Fan, Y. Liu, L. Wang and T. Zhang, Sens. Actuators, B, 2011, 160, 713–719.
38. F. Qu, C. Feng, C. Li, W. Li, S. Wen, S. Ruan and H. Zhang, Int. J. Appl. Ceram. Technol., 2014, 11, 619–625.
39. G. Korotcenkov, I. Boris, V. Brinzari, S. H. Han and B. K. Cho, Sens. Actuators, B, 2013, 182, 112–124.

---