Enhanced sensitivity and selectivity of brush-like SnO₂ nanowire/ZnO nanorod heterostructure based sensors for volatile organic compounds

Abstract

Brush-like SnO₂ nanowires have been grown by pulsed laser deposition on ZnO nanorods synthesized by the hydrothermal method. SnO₂ nanowire/ZnO nanorod heterostructures have been used for sensing several volatile organic compounds (VOCs). The heterostructure sensor exhibits higher response compared to that of control ZnO nanorods. The potential barriers formed at the SnO₂–ZnO and SnO₂–SnO₂ interface are proposed to be responsible for an improved sensing performance over the pure ZnO nanorods. The effect of the length of SnO₂ nanowires on the performance of triethylamine, toluene, ethanol, acetic acid, acetone, and methanol sensing has been studied. It is found that the response to the VOCs greatly depends on the length of the brush-like SnO₂ nanowires. The SnO₂/ZnO heterostructures can be successfully used to discriminate acetone from other VOCs.

---

1. Introduction

One of the most important environment related issues is the monitoring of air quality. It is essential for the detection of volatile organic compounds (VOCs) that are toxic, such as ketones, nitro- and amino compounds, aldehydes, and alcohols, in air. The VOCs have been identified to be the causes of a disease called building-related illness.^1–4 The sensing of VOCs is also important for early detection of diseases by breath analysis. The practically feasible devices for sensing VOCs should consist of a low cost sensing element with tunable properties. Semiconductor metal oxide can be an ideal candidate for sensing element because of its low cost, simple synthesis technique, high sensitivity, low limit of detection, and short response and recovery times.^5–10 However, the performance of these oxides is still limited due to poor selectivity towards various VOCs. The VOCs have quite similar molecular structure and elemental composition which increases the difficulty of distinguishing one from the other. Acetone is considered to be an important marker for the detection of diabetes using breath analysis. Previous studies^11–14 have shown that the sensitivity of acetone and ethanol are quite close to each other. So, it is important to choose proper sensing element which is able to distinguish acetone from other VOCs.

In order to improve the sensing performance, the semiconductor metal oxides are usually either doped with indium, antimony, etc., or their surface is modified with different metals or metal oxides, or mixed with another metal oxide to form a composite. Sensors fabricated with mixed oxide of SnO₂–TiO₂ doped with Ag ion showed excellent selectivity to different VOCs such as acetone, ethanol, formaldehyde, and methanol at different working temperatures.¹⁵ The SnO₂–ZnO composite thin film based sensor exhibited high selectivity to ethanol compared to that of pure ZnO and SnO₂ thin films.¹⁶ Mesoporous ZnO–SnO₂ nanofibers showed excellent sensing properties towards ethanol, such as high sensitivity, reproducibility, and fast response.¹⁷ The SnO₂/α-Fe₂O₃ hierarchical nanofibers exhibited higher response as well as shorter response and recovery times for ethanol and acetone compared to that of pure α-Fe₂O₃ and SnO₂ nanostructures.¹⁸

In this paper, we report the growth of brush-like SnO₂ nanowire/ZnO nanorod heterostructures and their sensing performance towards different VOCs such as triethylamine, toluene, ethanol, acetic acid, acetone, and methanol. The sensing characteristics of these heterostructures has been studied as a function of the length of the SnO₂ nanowires, and compared with those obtained using pure ZnO nanorods. An excellent selectivity to acetone over other VOCs has been achieved by varying the growth morphology. The sensing mechanism of brush-like SnO₂ nanowire/ZnO nanorod heterostructure based sensors are discussed in details in the light of homo- and hetero-junction potential barriers.

2. Experimental

2.1 Material synthesis

ZnO nanostructures were grown by hydrothermal method on copper substrates. The detailed experimental procedure is reported elsewhere.¹⁹ Briefly, ammonia (25%, 6 ml) was added to zinc chloride solution (120 ml, 0.1 M), and the reaction solution was poured into glass bottles, with substrates dipped vertically into the solution. The closed bottles containing the reaction solution were heated for 3.5 h at 95 °C, and then cooled down to reach room temperature. At room temperature, the substrates were taken out of the bottles and washed with de-ionized water. The substrates were then dried in air to obtain a thin layer of ZnO nanorods. SnO₂ nanostructures were deposited by pulsed laser deposition (PLD) on ZnO nanorods. For this, a SnO₂ target of high purity (99.999%) was placed 4 cm distant from the ZnO covered copper substrate, and a KrF excimer laser (Lambda Physik COMPEX, λ = 248 nm, τ = 25 ns) with energy density of ∼2 J cm⁻² was used to ablate the target. The deposition of SnO₂ was carried out at 100 °C in oxygen atmosphere at a pressure of 1 × 10⁻¹ mbar for 4, 6, and 8 minutes, with a repetition rate of 10 Hz. The samples were then annealed at 500 °C for 30 minutes in an argon atmosphere.

The morphology of the grown heterostructures were characterized by field emission scanning electron microscope (ZEISS SUPRA 40) fitted with an energy-dispersive X-ray analyzer, and transmission electron microscope (JEOL JEM-2100) operated at voltage 200 kV. The phase of the heterostructures was analyzed at a grazing incidence angle of 2° using X-ray diffraction (Philips X-Pert MRD) with Cu Kα radiation (45 kV, 40 mA).

2.2 Gas sensor device fabrication and characterization

For fabrication of gas sensing devices, three SnO₂/ZnO heterostructure samples with different deposition times were detached from the copper substrates separately under ultrasonic treatment through dispersion in acetone. The amount of detached samples and the volume of acetone were kept same in all cases. This solution containing SnO₂/ZnO heterostructures (only ZnO for the control device) was drop casted separately on Au-interdigitated electrodes, and then heated at 120 °C for 15 minutes to remove the solvent.

For gas sensing studies, the measurements were carried out in a reactor made of stainless steel having a base with temperature control, under dynamic flow of test gases. Different VOCs such as triethylamine, toluene, ethanol, acetic acid, acetone, and methanol were used as test gases, with air as a carrier gas. The flow rates of these gases were controlled by mass flow controllers (MFCs) to obtain the desired test gas concentration. The sensor elements were electrically contacted with the help of two probes attached to the reactor. The resistance of the sensor elements was measured by an Agilent 34972A LXI Data Acquisition (DAQ) unit equipped with digital multimeter and 34901A 20 channel multiplexer switches. The data acquisition was carried out using BenchLink Data Logger Pro software. A four channel mass flow controller power supply (MKS 247) was used to control the mass flow controllers.

3. Results and discussion

3.1 Material characterization

The FESEM images of the ZnO and SnO₂/ZnO heterostructures over the interdigitated electrodes are shown in Fig. 1. The micrographs reveal that a high density of hexagonal nanorods is oriented randomly over the electrode. The nanorods are 1–2 μm long having a diameter of ∼250–350 nm. The surface of pure ZnO nanorods is quite smooth, as shown in the inset of Fig. 1(a). For SnO₂/ZnO heterostructures, shown in Fig. 1(b)–(d), it is observed that SnO₂ is conformally covered on the surface of ZnO nanorods. The surface of SnO₂ is relatively rough as compared to pure ZnO nanorods.

Fig. 1 FESEM images of the (a) ZnO nanostructures; and SnO₂/ZnO heterostructures, with deposition time of SnO₂ of (b) 4, (c) 6, and (d) 8 minutes.

In order to have a closer view of these SnO₂/ZnO heterostructures, TEM analysis has been performed on these samples. Fig. 2(a)–(c) show the TEM images of SnO₂/ZnO heterostructures, with SnO₂ deposition time of 4, 6, and 8 minutes, respectively. Interestingly, SnO₂ forms a brush-like morphology on the surface of ZnO nanorods, instead of forming a continuous layer. It appears that high density of SnO₂ nanowires emerge out from the surface of the ZnO nanorods covering them uniformly. The length of the SnO₂ nanowires increases with increase in deposition time, and is found to be ∼25, 40, and 55 nm long for deposition time of 4, 6, and 8 minutes, respectively (Fig. 2(a)–(c)). A magnified view of the squared portion of Fig. 2(c) is presented in Fig. 2(d), along with its EDX spectra in the inset. As seen from Fig. 2(d), the density of the nanowires is not same at all places over the nanorod, making the surface of the nanorods more rough and porous. The EDX analysis reveals the presence of Zn, Sn, and O elements. The deposited SnO₂/ZnO heterostructures, with SnO₂ deposition time of 4, 6, and 8 minutes, are hereafter referred to as ZSO1, ZSO2, and ZSO3 samples, respectively.

Fig. 2 TEM images of the SnO₂/ZnO heterostructures, with deposition time of SnO₂ of (a) 4, (b) 6, and (c) 8 minutes. (d) Magnified view of the squared portion in part (c). Inset shows the corresponding EDX spectra. The dashed lines in parts (a), (b), and (c) are intended to guide the eye.

X-ray diffraction spectra of nanorods with and without SnO₂ nanowires are presented in Fig. 3. The diffraction peaks of pure ZnO are sharp and correspond to wurtzite ZnO structure. In ZSO1, diffraction peaks of SnO₂ phase are absent because of its relatively low fraction compared to ZnO. However, the signals corresponding to wurtzite ZnO and tetragonal rutile SnO₂ phases are distinguishably present in ZSO2 and ZSO3 samples. Diffraction peaks corresponding to secondary phases such as ZnSnO₃, Zn₂SnO₄, and SnO are absent, which confirm that the heterostructures are composed of ZnO and SnO₂ phases only.

Fig. 3 XRD pattern of (a) ZnO nanorods, (b) ZSO1, (c) ZSO2, and (d) ZSO3 heterostructures.

3.2 Gas sensing characteristics

The sensing performance of control ZnO and SnO₂/ZnO heterostructures have been investigated for different VOCs such as triethylamine, toluene, ethanol, acetic acid, acetone, and methanol. The response (S) of the sensor elements has been estimated using the relation,where R_a and R_g are the measured resistance of the sensor in air and test gas, respectively.

In order to find out the optimum working temperature for VOC sensing, the response of SnO₂/ZnO heterostructure based sensors as a function of temperature is presented in Fig. 4. The response of the sensors for all the VOCs except methanol increases with rise of temperature upto 300 °C, and then decreases with further increase of temperature. For methanol, the response of the sensors is found to be highest at 250 °C. The reason for this lies in the dynamic equilibrium attained between the adsorption and desorption processes with the rise of temperature.²⁰ The adsorption of VOCs increases initially and attains equilibrium at a certain temperature. With further increase of temperature, desorption dominates over absorption, leading to a decrease in the gas sensitivity. So, further measurements have been carried out at the corresponding optimum working temperatures.

Fig. 4 Response of sensor based on (a) ZSO2 to 115 ppm triethylamine, (b) ZSO2 to 29 ppm toluene, (c) ZSO1 to 57 ppm ethanol, (d) ZSO2 to 115 ppm acetic acid, (e) ZSO2 to 231 ppm acetone, and (f) ZSO2 to 115 ppm methanol, as a function of working temperature.

The effect of brush-like SnO₂ nanowires on the sensing performance has been examined by measuring the transients of pure ZnO nanorods and SnO₂/ZnO heterostructures to three representative VOCs viz. toluene, ethanol, and acetone. The sensing transient to toluene, ethanol, and acetone, measured at an operating temperature of 300 °C, is shown in Fig. 5(a)–(c), respectively. From Fig. 5(a), it is revealed that the response for 14–115 ppm toluene ranges from 2 to 4.8 for ZnO nanorods and increases for ZSO1 sensor ranging from 2.5 to 6.3. But the response is found to decrease for ZSO2 and ZSO3 based sensors, in fact it is lower than that exhibited by control ZnO. Thus, ZSO1 sensors show the highest response, which is ∼1.3 times higher than that of ZnO for 14–115 ppm toluene. Fig. 5(b) shows the ethanol sensing characteristics using the ZnO nanorods and SnO₂/ZnO heterostructures. The ZnO nanorod sensor exhibits a response ranging from 3.2 to 5.3 for 29–230 ppm ethanol. The response exhibited by ZSO1 is almost comparable to that of ZnO, but is found to increase for ZSO2 and ZSO3 based sensors. The highest response is obtained for ZSO3 and ranges from 5.9 to 16 for 29–230 ppm ethanol, which is ∼1.8 to 3 times higher than that of control ZnO. The acetone sensing transients are shown in Fig. 5(c). The response exhibited by ZnO sensors for 115–693 ppm acetone ranges from 4.2 to 16.6. The SnO₂/ZnO heterostructure sensors show higher response than that of pure ZnO. The highest response is observed for ZSO2 based sensor ranging from 21.4 to 48.4 for 115–693 ppm acetone, which is ∼5.1 to 3 times higher than that of control sample. Thus, interestingly, the sensing behavior of the SnO₂/ZnO heterostructures towards three VOCs is found to be different from each other, depending on the length of the SnO₂ nanowire brushes.

Fig. 5 Sensing responses of ZnO nanorods and SnO₂/ZnO heterostructure based sensors to (a) toluene, (b) ethanol, and (c) acetone, at 300 °C.

To investigate the selectivity of sensing of VOCs, the performances of all the sensors to 115 ppm triethylamine, toluene, ethanol, acetic acid, acetone, and methanol are presented in Fig. 6. The response of pure ZnO nanorods to all the VOCs except methanol is quite close to each other, indicating the lack of selectivity of ZnO nanorod sensors for the studied VOCs. For ZSO1 sensor, the response for ethanol, acetone, and methanol remains almost the same, and that of triethylamine, toluene, and acetic acid increases slightly compared to that of ZnO nanorods. However, the difference in responses to the VOCs is not so prominent indicating that there is lack of selectivity with this sensor as well. The ZSO2 sensor shows highest response for acetone (21.4), in comparison to the other VOCs (2.1–7.5) for a constant concentration of 115 ppm. Thus, the response for acetone is higher by a factor of three or more than that for the other VOCs. The result shows that ZSO2 heterostructures can be used for the selective detection of acetone. The ZSO3 based sensor shows highest response to ethanol (12.7), which is quite close to that observed for acetic acid (9.3), indicating the lack of selectivity with this sensor. In addition to the VOCs, the sensing performance of these sensors to different reducing gases such as CO and H₂ has also been investigated. The highest response is found to be ∼1.8 for 100 ppm CO and ∼1.2 for 500 ppm H₂ (results not presented here) for SnO₂/ZnO heterostructures, which are much lower compared to that of acetone. Therefore, SnO₂/ZnO heterostructures can be considered to be an excellent candidate for the selective detection of acetone over other VOCs by varying the morphology of the grown nanostructures.

Fig. 6 Sensing responses of ZnO nanorods and SnO₂/ZnO heterostructure based sensors to 115 ppm triethylamine, toluene, ethanol, acetic acid, acetone, and methanol.

The responses of ZSO1, ZSO2, and ZSO3 sensors as a function of concentration of toluene, acetone, and ethanol, respectively (since highest response has been obtained with the corresponding sensor) are shown in Fig. 7 as a representative plot. The response of a gas sensor based on semiconducting oxide is generally represented as,²¹

S = A_gP_g^β (2)

where A_g is a prefactor, P_g the partial pressure of VOC, and β is the exponent. Fitting the experimental data with eqn (2), the value of β were found to be 0.43 ± 0.06, 0.48 ± 0.06, and 0.44 ± 0.07 for toluene, ethanol, and acetone, respectively, which is close to the value of 0.5 possessed by an ideal microstructure²¹ based sensor. A small deviation of the value of ‘β’ from the ideal is attributed to the agglomeration or zones of the microstructure which are less sensitive to VOCs than others.²¹ The detection limit of toluene, acetone, and ethanol for sensors based on ZSO1, ZSO2, and ZSO3, respectively, is estimated by setting the criterion for minimum response detectable to be 1.2. The estimated limit of detection has been found to be approximately 2, 1, and <0.2 ppm for toluene, ethanol, and acetone, respectively. Thus, the SnO₂/ZnO heterostructure based sensors are capable of detecting quite low concentration of the appropriate VOCs.

Fig. 7 Response of SnO₂/ZnO heterostructure based sensors as a function of concentration of toluene, ethanol, and acetone. The solid line is fit to the experimental data.

3.3 Gas sensing mechanism

The sensing mechanism of the SnO₂/ZnO heterostructures can be explained using the space charge model²² in semiconductor junctions. Upon exposure to air, the metal oxides adsorb oxygen molecules on their surface, which in turn extract electrons from the conduction band of the metal oxides forming oxygen ions (O⁻, O²⁻, and O₂⁻). Thus a depletion layer is formed on the oxide surface which narrows the conduction channel resulting in a high resistance. The sensing mechanism is shown schematically in Fig. 8(a)–(d). When these oxides are exposed to VOCs such as triethylamine, toluene, acetic acid, and acetone, the oxygen ions will react with them according to the following reactions,^23–26

N(C₂H₅)₃ + O₂⁻ → H₂O + CO₂ + N₂ + e⁻ (3)

C₇H₈ + 2O⁻ ↔ H₂O + C₇H₆ − O + 2e⁻ (4)

CH₃COOH + 4O²⁻ → 2CO₂ + 2H₂O + 8e⁻ (5)

C₃H₆O + 8O⁻ → 3CO₂ + 3H₂O + 8e⁻ (6)

where eqn (3)–(6) are the reactions of the oxygen ions with triethylamine, toluene, acetic acid, and acetone, respectively. The decomposition reaction of ethanol is related to the acid–base properties of the metal oxide, and is given by,²⁷

C₂H₅OH → C₂H₄ + H₂O (acidic oxide) (7)

C₂H₅OH → CH₃CHO + H₂ (basic oxide) (8)

Fig. 8 Schematic diagram of sensing mechanism of (a and b) ZnO nanorods and (c–e) SnO₂/ZnO heterostructures. SnO₂ nanowires are shown only on the two sides of ZnO nanorods in parts (c) and (d) for clarity.

ZnO is known to be a basic oxide²⁸ and SnO₂ exhibits acidic as well basic properties.²⁷ For ZnO nanorods, only dehydrogenation to intermediate CH₃CHO is expected, whereas dehydration to intermediate C₂H₄ in addition to dehydrogenation will occur for SnO₂/ZnO heterostructures. The intermediate CH₃CHO and C₂H₄ can react with oxygen ions to produce CO₂ and H₂O (ref. 29) following,

CH₃CHO + 5O⁻ → 2CO₂ + 2H₂O + 5e⁻ (9)

and

C₂H₄ + 6O⁻ → 2CO₂ + 2H₂O + 6e⁻ (10)

On the other hand, methanol reacts with the oxygen ions producing either formaldehyde or formic acid and water³⁰ following,

CH₃OH + O⁻ → HCOH + H₂O + e⁻ (11)

CH₃OH + O₂⁻ → HCOOH + H₂O + e⁻ (12)

The released electrons, obtained from the reaction of oxygen ions with the VOCs, will return to the conduction band of the oxide, and thereby decreases the width of the depletion layer by widening the conduction channel. This leads to decrease in the resistance.

Another mechanism that may modulate the electron transport properties of the metal oxides is the formation of hetero- and homo-junctions at the interfaces. The electron transport in the SnO₂/ZnO heterostructures is different from that of ZnO nanorods due to the formation of multiple junctions at the interfaces. As observed from the TEM images of Fig. 2, the SnO₂ nanowires are in contact with one another due to their high density. So, a homo-junction potential barrier is formed at the interface of the nanowires. When SnO₂ and ZnO forms a hetero-junction, electrons will flow from SnO₂ (work function of 4.9 eV)³¹ to ZnO (with relatively higher work function of 5.2 eV)³² till their Fermi levels equilibrate. Thus an electron depletion layer is formed at the interface which bends the energy band, as shown in Fig. 8(e). The built-in potential at the interface of ZnO and SnO₂ is given by the difference between their work functions, i.e. 0.3 eV. If Φ₁ and Φ₂ are the homo- and hetero-junction potential barriers, respectively, the contributions of both the barriers to the modulation of electron transport can be given by that of an effective potential barrier, Φ_eff. Then the conductivity can be expressed as,³³

G = G₀exp(−Φ_eff/k_BT) (13)

where G₀ is a constant parameter, k_B the Boltzmann's constant, and T is absolute temperature. In air, the electrons are trapped in the metal oxides due to the adsorbed oxygen molecules, which will increase the height and width of Φ_eff. When the metal oxides are exposed to VOCs, the trapped electrons will return back to the conduction band, which leads to the decrease in Φ_eff.^34,35 Therefore, homo- and hetero-junction potential barriers with controllable height and width modulate the electron transport, and accordingly play an important role in controlling the sensing characteristics of the SnO₂/ZnO heterostructures leading to better sensing performance than that of the ZnO nanorods only.

The sensing performance of SnO₂/ZnO heterostructures, shown in Fig. 6, reveals that the response to the VOCs is strongly dependent on the length of the SnO₂ nanowires. This may be related to the available inter-nanorod spacing with the change of the length of SnO₂ nanowires with respect to the molecular size of the VOCs. In ZSO1 sensor, the length of SnO₂ nanowires is lowest and the spacing between the nanorods is quite large. So, relatively smaller fraction of the injected VOCs is adsorbed on nanowire surfaces, leading to lower sensitivity without exhibiting any selectivity. The spacing between the nanorods in ZSO2 is less compared to that of ZSO1 due to the presence of relatively longer SnO₂ nanowires. So, a comparatively larger fraction of the injected acetone, acetic acid, methanol, and ethanol molecules are adsorbed on nanowire surfaces, which will react with the oxygen ions resulting in a higher response than that obtained in ZSO1. The size of an ethanol, acetic acid, and methanol molecule (kinetic diameter of 0.45, 0.436, and 0.36 nm,^36,37 respectively) is smaller than that of acetone (kinetic diameter of 0.469 nm (ref. 37)), allowing more acetone molecules to be absorbed. This leads to a higher response for 115 ppm acetone than ethanol, acetic acid, and methanol in ZSO2. Since the nanorods are oriented randomly over the interdigitated electrode, the spacing between the nanorods becomes narrower with the increase of the length of the SnO₂ nanowires, thereby reducing the diffusion of larger sized triethylamine and toluene molecules (kinetic diameter of 0.78 and 0.585 nm,^38,39 respectively). Thus a portion of the heterostructure remains unexposed to them, leading to a decrease in the sensitivity of triethylamine and toluene in ZSO2 compared to that of ZSO1. Further increase of the length of the SnO₂ nanowires lowers the effective spacing between the nanorods in ZSO3. This leads to an increase in response for ethanol, acetic acid, and methanol sensing compared to that of ZSO2. Since the size of the acetone, triethylamine, and toluene molecules is higher than that of ethanol, acetic acid, and methanol, so a larger portion of the heterostructures remain unexposed to them, resulting in a decrease of their sensitivity. Therefore, different spacing between the nanorods with the change of the length of SnO₂ nanowires and the size of the VOC molecules result in selectivity with changing microstructures. However, further investigations are required to draw a firm conclusion.

4. Conclusions

We have successfully grown brush-like SnO₂ nanowire/ZnO nanorod heterostructures, with varying length of the nanowires. The potential of these heterostructures as VOC sensors have been tested in terms of triethylamine, toluene, ethanol, acetic acid, acetone, and methanol. The response is found to depend strongly on the length of the SnO₂ nanowires. The highest response was observed for SnO₂/ZnO heterostructures for a certain length of SnO₂ nanowires, which was higher than that obtained from pure ZnO nanorods for all the VOCs. The SnO₂/ZnO heterostructures with 40 nm long SnO₂ nanowires showed a high selectivity to acetone over other VOCs. The estimated limit of detection for acetone has been found to be approximately <0.2 ppm for ZSO2 sensor. The mechanism for improved sensing and selectivity has been discussed in details.

Acknowledgements

This work was supported by DST-GPU project. One of the authors (S. Santra) acknowledges the Department of Science and Technology (DST), India for the support of the work (project no. SR/S2/RJN-104/2011).

References

1. J. W. Gardner and W. E. Gardner, Insight, 1997, 39, 865–869.
2. M. Odlyha, G. M. Foster, N. S. Cohen, C. Sitwell and L. Bullock, J. Environ. Monit., 2000, 2, 127–131.
3. C. Arnold, M. Harms and J. Goschnick, IEEE Sens. J., 2002, 2, 179–188.
4. S. Zampolli, I. Elmi, F. Ahmed, M. Passini, G. C. Cardinali, S. Nicoletti and L. Dori, Sens. Actuators, B, 2004, 101, 39–46.
5. T. Itoh, J. Z. Wang, I. Matsubara, W. Shin, N. Izu, M. Nishibori and N. Murayama, Mater. Lett., 2008, 62, 3021–3023.
6. C. Q. Ge, C. S. Xie and S. Z. Cai, Mater. Sci. Eng., B, 2007, 137, 53–58.
7. B. L. Zhu, C. S. Xie, J. Wu, D. W. Zeng, A. H. Wang and X. Z. Zhao, Mater. Chem. Phys., 2006, 96, 459–465.
8. Y. S. Kim, Sens. Actuators, B, 2009, 137, 297–304.
9. M. A. El Khakani, R. Dolbec, A. M. Serventi, M. C. Horrillo, M. Trudeau, R. G. Saint-Jacques, D. G. Rickerby and I. Sayago, Sens. Actuators, B, 2001, 77, 383–388.
10. C. de Julián Fernández, M. G. Manera, G. Pellegrini, M. Bersani, G. Mattei, R. Rella, L. Vasanelli and P. Mazzoldi, Sens. Actuators, B, 2008, 130, 531–537.
11. B. L. Zhu, C. S. Xie, W. Y. Wang, K. J. Huang and J. H. Hu, Mater. Lett., 2004, 58, 624–629.
12. H. Gong, Y. J. Wang, S. C. Teo and L. Huang, Sens. Actuators, B, 1999, 54, 232–235.
13. Z. Jing and S. Wu, Mater. Lett., 2006, 60, 952–956.
14. Z. Jie, H. Li-Hua, G. Shan, Z. Hui and Z. Jing-Gui, Sens. Actuators, B, 2006, 115, 460–464.
15. Z. Wen and L. Tian-mo, Phys. B, 2010, 405, 1345–1348.
16. 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, 318–324.
17. X. Song, Z. Wang, Y. Liu, C. Wang and L. Li, Nanotechnology, 2009, 20, 075501-1–075501-5.
18. B. B. Wang, X. X. Fu, F. Liu, S. L. Shi, J. P. Cheng and X. B. Zhang, J. Alloys Compd., 2014, 587, 82–89.
19. T. Rakshit, S. Mandal, P. Mishra, A. Dhar, I. Manna and S. K. Ray, J. Nanosci. Nanotechnol., 2012, 12, 308–315.
20. J. Xu, X. Jia, X. Lou, G. Xi, J. Han and Q. Gao, Sens. Actuators, B, 2007, 120, 694–699.
21. R. W. J. Scott, S. M. Yang, G. Chabanis, N. Coombs, D. E. Williams and G. A. Ozin, Adv. Mater., 2001, 13, 1468–1472.
22. C. Xu, J. Tamaki, N. Miura and N. Yamazoe, Sens. Actuators, B, 1991, 3, 147–155.
23. B. Liu, L. Zhang, H. Zhao, Y. Chen and H. Yang, Sens. Actuators, B, 2012, 173, 643–651.
24. Y. Zeng, T. Zhang, L. Wang, M. Kang, H. Fan, R. Wang and Y. He, Sens. Actuators, B, 2009, 140, 73–78.
25. C. Wang, S. Ma, A. Sun, R. Qin, F. Yang, X. Li, F. Li and X. Yang, Sens. Actuators, B, 2014, 193, 326–333.
26. N. H. Al-Hardan, M. J. Abdullah and A. Abdul Aziz, Appl. Surf. Sci., 2013, 270, 480–485.
27. H.-W. Cheong and M.-J. Lee, J. Ceram. Process. Res., 2006, 7, 183–191.
28. X. Ma, H. Song and C. Guan, Sens. Actuators, B, 2013, 188, 193–199.
29. S. Shi, Y. Liu, Y. Chen, J. Zhang, Y. Wang and T. Wang, Sens. Actuators, B, 2009, 140, 426–431.
30. N. G. Patel, P. D. Patel and V. S. Vaishnav, Sens. Actuators, B, 2003, 96, 180–189.
31. R. Könenkamp, R. C. Word and M. Godinez, Nanotechnology, 2006, 17, 1858–1861.
32. Y. Liu, M. Zhong, G. Shan, Y. Li, B. Huang and G. Yang, J. Phys. Chem. B, 2008, 112, 6484–6489.
33. T. Weis, R. Lipperheide, U. Wille and S. Brehme, J. Appl. Phys., 2002, 92, 1411–1418.
34. V. V. Sysoev, J. Goschnick, T. Schneider, E. Strelcov and A. Kolmakov, Nano Lett., 2007, 7, 3182–3188.
35. A. Kolmakov, D. O. Klenov, Y. Lilach, S. Stemmer and M. Moskovits, Nano Lett., 2005, 5, 667–673.
36. T. Borjigin, F. Sun, J. Zhang, K. Cai, H. Ren and G. Zhu, Chem. Commun., 2012, 48, 7613–7615.
37. T. C. Bowen, R. D. Noble and J. L. Falconer, J. Membr. Sci., 2004, 245, 1–33.
38. R. Szostak, in Handbook of Molecular Sieves, Van Nostrand Reinhold, New York, 1992, p. 184.
39. C. D. Baertsch, H. H. Funke, J. L. Falconer and R. D. Noble, J. Phys. Chem., 1996, 100, 7676–7679.

---

Footnote

† Present address: IIT Kanpur, 208 016, India.

---