Enhanced LPG sensing-performance at room temperature of poly(o-anisidine)–CeO₂ nanocomposites

Abstract

We report a fast-responding and selective room temperature liquefied petroleum gas (LPG) sensor based on poly(o-anisidine)–cerium oxide (POA–CeO₂) nanocomposites. The POA–CeO₂ nanocomposites were prepared by an in situ chemical oxidative polymerization of o-anisidine in the presence of pre-prepared CeO₂ nanoparticles (size ∼5–10 nm). The pre-prepared CeO₂ nanoparticles were synthesized by a simple and rapid microwave-assisted method. The POA, CeO₂ nanoparticles and their nanocomposites were characterized by X-ray diffraction (XRD), Fourier transform infrared (FTIR) spectroscopy and transmission electron microscopy (TEM). The results of the characterization revealed that interaction occurred between the POA and CeO₂ nanoparticles, leading to the formation of POA–CeO₂ nanocomposites. The LPG sensing-performance of the nanocomposites was investigated and compared with that of sensors based on POA. The sensor based on a POA–CeO₂ nanocomposite with known weight percentage of CeO₂ showed better LPG sensing properties than that of pure POA. Upon exposure to 3.4% LPG, the nanocomposite with 30 wt% CeO₂ exhibited an enhanced response (∼2.06%), very fast response (∼3–4 s), full recovery within 20 s in air and, most notably, better selectivity at room temperature. Furthermore, the lower detection limit was ∼0.09% which is lower than the permissible explosive limit for LPG. The reproducibility and stability study reveals that these nanocomposites are highly stable. Thus, the combination of the CeO₂ nanoparticles and POA offers an attractive selective sensing-performance for LPG at room temperature with fast response/recovery rate.

---

1. Introduction

Liquefied Petroleum Gas (LPG) is highly inflammable and colourless. It is widely used as fuel in domestic and industrial fields and popularly known as cooking gas.^1–4 The main constituents of LPG are butane (70% to 80%), propane (5% to 10%) and propylene, butylene, ethylene and methane (1% to 5%).⁵ It is generally mixed with compounds of sulphur (methyl mercaptan and ethyl mercaptan) to give a foul smell, so that its leakage can be noticed easily.⁶ The permissible exposure limit (PEL) for LPG as specified by the National Institute for Occupational Safety and Health (NIOSH) and Occupational Safety and Health Administration (OSHA) standards is 1000 ppm (0.1%).⁷ It is potentially hazardous due to the high possibility of accidental explosions caused by leakage or by human error. Hence, it is crucial to detect LPG in its early stages to raise the alarm and perform effective suppression. This has stimulated considerable interest for scientific research to develop reliable, efficient, simple and cost-effective chemical sensors to monitor LPG selectively with good sensitivity. Metal oxides^5,8–13 and conducting polymers have been used as sensing elements for various LPG sensing applications.¹⁴ However, metal oxide-based sensors work efficiently at high operating temperatures, typically in the range of 200–500 °C, which increases power consumption, reduces sensor life and limits the portability.^15,16 This disadvantage impedes their practical usage in many real time applications. Over the last decade, considerable research has been devoted towards the synthesis of novel sensing materials having enhanced performance at room temperature.

Conducting polymers are another important class of low-cost sensing material demonstrated to have rapid response and room temperature operation, which is a major advantage for gas detection, especially for LPG.^17–21 Among various conducting polymers, polyaniline (PANI) has been recognized as an interesting candidate in sensor applications due to its easy synthesis, unique electrical properties, relatively good environmental stability, sensitivity to some volatile organic compounds and ability to be attached onto flexible film.^17,18,22 However, it still exhibits certain limitations such as a low gas response, irreversibility, long recovery time and lack of specificity. Although gas-sensing performances of different conducting polymers for H₂,²³ CO,²⁴ NH₃,²⁵ H₂S,²⁶ NO₂ (ref. 27) etc. have been studied, hardly any attempt has been made to investigate their LPG sensing properties. More recently, Birajadar et al.¹⁴ studied the room temperature LPG sensing behavior of PANI thin films synthesized by a cost-effective oxidative polymerization technique. The PANI thin film showed the highest response of 22 at 300 ppm with response and recovery times of 42 and 80 s, respectively. However, the other aspects of the LPG sensing such as sensing selectivity, reproducibility and stability have not been studied by those authors.

Among several different strategies for fabricating selective and room temperature operative gas sensors, the combination of nanostructured metal oxides with conducting polymers has been particularly interesting.^28,29 Combination of conducting polymers and nanostructured metal oxides results in the generation of new gas-sensing materials having drastically improved gas-sensing performance for various gases compared with their individual components.^30–33 Such nanocomposites have shown excellent sensing performance under ambient conditions and therefore can be used for the fabrication of low-cost room temperature gas sensors. In recent years, conducting polymer–metal oxide nanocomposites such as POA–SnO₂,³⁴ polythiophene–SnO₂,³⁵ polypyrrole–zinc oxide,³⁶ poly(o-chloroaniline)–ZnO³⁷ and polyaniline–ZnO⁷ have been synthesized and their LPG sensing properties have also been investigated by a few research groups.

Cerium oxide (CeO₂) is a technologically important n-type semiconductor and has been used to detect CO,³⁸ acetone and NO³⁹ in an oxygen environment at operating temperatures in the range of 300–400 °C. However, hardly any attempt has been made to investigate the LPG sensing properties of CeO₂ nanoparticles. Before this study, there are only a few reports dealing with the synthesis of PANI–CeO₂ composites for NH₃ (ref. 40) and LPG⁴¹ sensing. Wang et al.⁴⁰ recently reported the synthesis of core–shell CeO₂ nanoparticles@cross-linked PANI nanohybrids and investigated their NH₃ sensing properties. When the ratio of CeO₂ to aniline was 4, the nanohybrid showed an enhanced response of ∼6.5 to 50 ppm of NH₃ at room temperature with a shorter detection time of ∼57.6 s than pure PANI (response of 2.6 and 290 s of response time). Khened et al.⁴¹ reported the synthesis and LPG sensing behavior of PANI–CeO₂ composites at room temperature. The composite with 20 wt% of CeO₂ showed a maximum sensitivity to LPG at room temperature. However, the LPG sensing properties of the pure PANI have not been investigated and compared with those of the composites. Also, the other aspects of the gas-sensing such as response and recovery time, reproducibility and stability have not been studied by those authors.

The present study reports an LPG sensor featuring a POA–CeO₂ nanocomposite as the sensitive layer. To the best of our knowledge, the LPG sensing characteristics of a POA–CeO₂ nanocomposite at room temperature have not been investigated as yet. In this study, the POA–CeO₂ nanocomposite was synthesized using an in situ chemical oxidative polymerization method and its LPG sensing performance was investigated at room temperature. The POA was chosen as an organic counterpart for this study to explore the possibility of utilizing it as an alternative to PANI for gas-sensing applications. The fabricated POA–CeO₂ (30 wt%) nanocomposite sensor shows enhanced LPG response at room temperature, excellent reproducibility, rapid response, fast recovery, good sensing selectivity, stability and lower detection limit (∼0.09% LPG < PEL). A sensing mechanism is also discussed based on experimental findings.

2. Experimental

2.1 Materials

The purity of all chemicals was analytical grade and they were used as-received, without any further purification except o-anisidine. Ammonium persulphate ((NH₄)₂S₂O₈), hydrochloric acid (HCl), ammonium cerium(IV) nitrate ((NH₄)₂Ce(NO₃)₆), propylene glycol (C₃H₈O₂), liquid ammonia (30%) and o-anisidine were procured from Qualigens Fine Chemicals, India. o-Anisidine was doubly distilled prior to its use in the synthesis.

2.2 Preparation of CeO₂ nanoparticles

CeO₂ nanoparticles were prepared using a microwave-assisted route following the earlier work reported by our group.⁴² Briefly, 0.1 M each of ammonium cerium(IV) nitrate and propylene glycol were dissolved in double-distilled water and stirred continuously for 1 h at room temperature (27 °C). Ammonia was added drop-wise to this solution with continuous stirring until the final pH value of approximately 10 was achieved in the solution. The resulting pale-yellow-colored precipitate was filtered, washed several times with double-distilled water and alcohol, and subjected to microwave irradiation at 600 W microwave power for 10 minutes. Microwave irradiation was performed with a domestic microwave oven (CQ138S, Samsung, Japan) operated at a microwave frequency of 2.45 GHz with a maximum power of 800 W. After irradiation, the yellow-colored precipitate was harvested by centrifugation, washed several times using double-distilled water and ethanol, and then dried in an oven at 90 °C overnight to obtain the end-product for further characterization.

2.3 Preparation of POA–CeO₂ nanocomposites

The POA–CeO₂ nanocomposites were prepared through chemical polymerization of o-anisidine at 0–5 °C in air in the presence of pre-prepared CeO₂ nanoparticles. In a typical synthesis, CeO₂ nanoparticles were dispersed into 1 M HCl solution under ultrasonication for 2 h. After 2 h of ultrasonication, o-anisidine was introduced into the above solution under continuous stirring. Then, ammonium persulphate (0.1 M) was slowly added drop-wise to the well-dispersed reaction mixture for 2 h with continuous stirring at 0–5 °C for another 24 h. The dark green-colored precipitate produced in the reaction was harvested by filtration and washed repeatedly with methanol and double-distilled water. Finally, the filtered POA–CeO₂ nanocomposite was dried at 80 °C under vacuum for 24 h. Different POA–CeO₂ nanocomposites were prepared using 10, 20, 30, 40 and 50 wt% of CeO₂ nanoparticles with respect to o-anisidine, which are referred to as POA–CeO₂-10, POA–CeO₂-20, POA–CeO₂-30, POA–CeO₂-40 and POA–CeO₂-50, respectively. For comparison, pure POA was also synthesized following the same procedure (without the use of CeO₂ nanoparticles), characterized and employed as a sensing material for LPG.

2.4 Characterization

X-ray powder diffraction (XRD) analysis was performed on a Bruker diffractometer (D8, Advance, Bruker AXS model) with Cu Kα monochromatic radiation (λ = 1.5406 Å) at 40 kV and 40 mA in the range of 5–80° (2θ). The surface morphology and microstructure of the products were observed using a transmission electron microscope (TEM, Tecnai G² U20, FEI, USA) operating at an accelerating voltage of 200 kV. FTIR spectra of the samples were measured using the KBr pellet method on a Shimadzu FTIR spectrophotometer (FTIR-4800, Shimadzu, Japan) in the spectral range 4000–400 cm⁻¹.

2.5 LPG sensing measurements

In this work, pellets (diameter ∼1 cm and thickness ∼0.1 cm) of the POA and POA–CeO₂ nanocomposite powders (0.2 g) were formed using a hydraulic press machine (Techno Search Instruments, Mumbai, India) under a uniaxial pressure of 15 MPa at room temperature. The pellets were dried in an oven at 90 °C for 3 h and the electrical contact leads were fixed 0.7 cm apart with the help of silver paste on the surface of the pellet to form the sensing element. The LPG sensing experiments were carried out in a static gas chamber in ambient air. The sensing element was kept directly on a heater in the chamber and the temperature was varied from room temperature to 100 °C. The temperature of the sensing element was monitored by chromel-alumel thermocouple placed in contact with the sensor. The known volume of the LPG was introduced into the gas chamber pre-filled with air and it was maintained at atmospheric pressure. The LPG was detected by monitoring and recording the change of electric current passing through the sensing element using an electrochemical measurement system (1287, Solartron, U.K.) controlled by CorrWare software from Scribner Associates Inc. supplied by Solartron, U.K. A constant dc voltage of 1 V was applied to the electrodes. The resistance of the sensing element as a function of time was calculated from this data using the software. The performance of the sensing element is presented in terms of the response (S), which is defined by the following relation:where R_a and R_g are values of the resistance of the sensing element in air and in the presence of LPG, respectively.

3. Results and discussion

3.1 Structural and morphological characteristics

The XRD patterns of the as-prepared CeO₂ nanoparticles and the POA–CeO₂-30 nanocomposite are shown in Fig. 1. The XRD pattern of the as-prepared CeO₂ nanoparticles [Fig. 1(a)] is in good agreement with that of the pure CeO₂ crystalline phase (JCPDS no. 81-0792). All of the diffraction peaks can be indexed to CeO₂ with a face-centered cubic structure. No other peaks were observed, indicating that no impurities were present and confirming that the adopted synthesis method gives pure CeO₂ nanoparticles. The average crystallite size was estimated by fitting the [1 1 1] diffraction peak (2θ = 27.76°) width using a Gaussian function and using the values of the angular position of the peak (θ) and peak full line width at half maximum (B) in the Debye–Scherrer equationwhere D is the average size of the crystallite, assuming that the grains are spherical and λ is the wavelength of the X-ray radiation. The average crystallite size calculated for the CeO₂ nanoparticles is in the range of 5–10 nm. The XRD pattern of POA [inset of Fig. 1] exhibits two characteristic diffraction peaks at 2θ values of 11.27° and 24.5° corresponding to (0 1 1) and (2 0 0) crystal planes of POA. These peaks are assigned to the scattering from POA chains at interplanar spacing and it indicates that pure POA has some degree of crystallinity. This is consistent with the results reported by Patil et al.³⁴

Fig. 1 XRD patterns of (a) CeO₂ nanoparticles and (b) POA–CeO₂-30 nanocomposite. Inset depicts the XRD pattern of POA.

As expected, the XRD pattern of the POA–CeO₂-30 nanocomposite [Fig. 1(b)] exhibits the characteristic diffraction peaks not only for the POA but also for the CeO₂ nanoparticles, which confirms the co-existence of CeO₂ nanoparticles with POA. It can be noted that the intensities of the diffraction peaks of POA are considerably reduced, suggesting that the CeO₂ nanoparticles hamper the crystallization of POA chains. Furthermore, the comparison of the XRD pattern of the nanocomposite with that of CeO₂ nanoparticles suggests that the crystal structure of CeO₂ was not affected by POA. Also, the intensities of the diffraction peaks for the nanocomposite are weaker than those for the pure CeO₂ nanoparticles, which is attributed to the deposition of POA on the surface of the nanoparticles.

The FTIR spectrum of CeO₂ nanoparticles (Fig. S1, ESI†) is similar to that reported in our earlier work.⁴² The main characteristic bands in the spectrum are assigned as follows; the bands at 538 and 851 cm⁻¹ correspond to Ce–O stretching vibrations,⁴³ the bands at 1061 and 1385 cm⁻¹ are associated with the characteristic vibrations of CeO₂ (ref. 44) and a broad band at ∼3320 cm⁻¹ is assigned to the Ce–OH stretching vibration. Fig. 2(a) and (b) show the FTIR spectra of POA and POA–CeO₂-30 nanocomposite. The FTIR spectrum of pure POA [Fig. 2(a)] is also consistent with the earlier work reported by our group^45,46 having strong bands at around 1568 and 1490 cm⁻¹ corresponding to quinoid (Q) rings (C=N stretching mode) and the benzoid (B) rings (C=C stretching mode), respectively, and the remaining bands are assigned as follows: a broad and weak band at ∼3436 cm⁻¹ is due to the characteristic N–H stretching vibration, the bands at ∼1254 and 1014 cm⁻¹ are assigned to the presence of an o-methoxy group in the POA, a band at ∼1207 cm⁻¹ is a measure of the degree of delocalization of electrons on POA, a band at ∼1123 cm⁻¹ is attributed to a plane bending vibration of C–H and a band at ∼805 cm⁻¹ indicates the presence of an ortho-substituted benzene ring.

Fig. 2 FTIR spectra of (a) POA and (b) POA–CeO₂-30 nanocomposite.

As expected, the FTIR spectrum of the POA–CeO₂-30 nanocomposite [Fig. 2(b)] shows the presence of characteristic bands attributable to both POA (1584, 1502, 1284, 1209, 1126, 1022, 596 and 457 cm⁻¹)^45,46 and CeO₂ nanoparticles (950 and 839 cm⁻¹).^42–44 However, the incorporation of CeO₂ nanoparticles leads to the shift of FTIR bands of POA to higher wavenumbers, which suggests that there is strong interaction between the POA and CeO₂ nanoparticles. Moreover, the ratio of intensities of quinoid and the benzoid bands calculated by integrating the corresponding band area were found to be ∼0.39 and 0.78 for POA and POA–CeO₂-30 nanocomposite, respectively. The higher value of for the nanocomposite indicates the increase in the degree of oxidation in POA due to the incorporation of CeO₂ nanoparticles.⁴⁷ More careful examination reveals that the intensity of the FTIR spectrum is observed to be significantly lower than that for POA. This may possibly be due to the fact that the CeO₂ nanoparticles present in the reaction mixture act as template areas for the growth of POA. The o-anisidine monomer gets adsorbed on the CeO₂ nanoparticles and polymerization proceeds on the surface of oxide particles when the oxidant (ammonium persulphate) is added to the reaction mixture. The polymerization of o-anisidine is an auto-accelerated process. That is, the conducting polymer formation is preferred in the region where some polymer has already been produced. Thus, adsorption of polymer onto the CeO₂ nanoparticles results in constrained chain growth around the nanoparticles and restricts the amplitudes of modes of vibration in POA, which in turn leads to the decrease in the intensities of bands in the FTIR spectrum.^48,49 Thus, the FTIR spectroscopy result confirms that the nanocomposite has been formed during chemical polymerization of o-anisidine in the presence of CeO₂ nanoparticles.

The TEM image of the CeO₂ nanoparticles [Fig. 3(a)] indicates the presence of CeO₂ nanoparticles with clear boundaries and crystalline with an average size of ∼10–12 nm. The corresponding selected area electron diffraction (SAED) pattern (Fig. 3(b)) further confirms the random orientations of the CeO₂ nanoparticles and there is no secondary phase. The TEM image of the POA–CeO₂-30 hybrid shown in Fig. 3(c) indicates the presence of dark colored CeO₂ nanoparticles are enwrapped and interconnected with POA in the nanocomposite. It is observed that the CeO₂ nanoparticles are homogeneously distributed in the hybrid. The SAED pattern of POA–CeO₂-30 hybrid [Fig. 3(d)] shows the diffraction rings composed of spots corresponding to the different planes of a face-centred cubic CeO₂, which is in agreement with the XRD result.

Fig. 3 TEM images of (a) CeO₂ nanoparticles and (b) POA–CeO₂-30 nanocomposite; (c) and (d) present the corresponding SAED patterns of CeO₂ nanoparticles and POA–CeO₂-30 nanocomposite, respectively.

3.2 LPG sensing characteristics

In order to demonstrate that the formation of a composite of POA with CeO₂ nanoparticles is an effective approach to enhance the gas-sensing properties of POA-based gas sensors, the sensors based on CeO₂ nanoparticles, POA and POA–CeO₂ nanocomposites were investigated for LPG detection at room temperature. The CeO₂ nanoparticles do not show sensor activity, while POA exhibits a response of ∼0.83% to 3.4% LPG at room temperature, thus, the components of the nanocomposite individually exhibit either poor response or no response to LPG. However, their nanocomposite shows sensor activity to 3.4% LPG at room temperature and it exhibits the properties of a p-type semiconductor, that is its resistance increases on exposure to reducing gas (e.g. LPG). This observation suggests that the LPG sensing mechanism of the POA–CeO₂ nanocomposite is governed by its POA counterpart, which is supported by the TEM study.

The response of the nanocomposite with 30 wt% CeO₂ towards 3.4% LPG at room temperature is about 2.5 times that of POA. The enhancement in the sensor functionality of the nanocomposite sensor may be attributed to the mutual interaction between POA and CeO₂ nanoparticles.^34,50,51 The two components of the nanocomposite, namely POA and CeO₂, are p-type and n-type semiconductors, respectively. The mutual interaction between them creates charge separation and generates the p–n heterojunctions in the nanocomposite. The R_a value of the POA–CeO₂-30 nanocomposite (∼333 kΩ) is ∼16 times higher than that of the POA sensor (∼21 kΩ). Thus, the significant increase of R_a is attributed to the extension of the electron depletion layer by the formation of p–n heterojunctions between the POA and CeO₂ nanoparticles.

The responses of the sensors based on POA–CeO₂ nanocomposites to 3.4% LPG at room temperature were further examined to determine the optimal content of CeO₂ nanoparticles in the nanocomposites. Fig. 4(a) shows the dependence of the LPG response on the CeO₂ content in the nanocomposite and the highest response is achieved for 30 wt% of CeO₂. The y-error bars represent the standard deviation at a fixed CeO₂ content and they are in the range of 0.06–0.10, indicating the low experimental variability. On the basis of this result, the nanocomposites for the sensing experiments in this study were synthesized with 30 wt% of CeO₂. It is well known that the gas response is greatly influenced by the operating temperature. Therefore, both the sensors were also investigated for LPG detection at different operating temperatures in the range between 25 °C (room temperature) and 100 °C. The results (Fig. S2(a), ESI†) reveal that the POA–CeO₂-30 nanocomposite sensor exhibits maximum LPG response ∼76% at 100 °C and the value is about 6.21 times higher than that of POA.

Fig. 4 (a) Bar chart showing the gas response of POA–CeO₂ nanocomposites with 0, 10, 20, 30, 40 and 50 wt% of CeO₂ nanoparticles to 3.4% LPG at room temperature; (b) bar chart showing the dependence of dielectric constant of the POA–CeO₂ nanocomposites on the content of CeO₂; (c) dynamic response–recovery characteristics of pure POA and POA–CeO₂-30 nanocomposite based sensors upon exposure to 3.4% LPG at room temperature and (d) repetitive responses of pure POA and POA–CeO₂-30 nanocomposite to 3.4% LPG at room temperature (25 °C). Error bar indicates the standard deviation in the measurement.

It is well-known that the charge transfer process is accompanied with the proportionate change in the dielectric property of the material. Consequently, the dielectric constant of the nanocomposite can be considered to be a measure of the formation of p–n heterojunctions. In order to substantiate the formation of p–n heterojunctions, the values of dielectric constant of the POA–CeO₂ nanocomposites were estimated at room temperature using the following relation^50,52

where C is the bulk capacitance, ε₀ is the permittivity of free space (=8.854 × 10⁻¹² F m⁻¹), A is cross sectional area and d is the thickness of the pellet. The value of C in eqn (3) is determined by using the relation⁵³where Z′ and Z′′ are the real and imaginary parts of the measured impedance and ω = 2πf is the angular frequency of the AC signal. The Nyquist impedance plots were recorded using a Solartron impedance gain phase analyzer (SI 1260, Solartron, U.K.) in a two-point configuration in the frequency range 40 Hz to 100 kHz with an alternating-current excitation potential of 10 mV. The dielectric constant of the nanocomposites as a function of CeO₂ content is shown in Fig. 4(b). The standard deviation in the measurement is in the range of 0.09–6.55.

It is seen that the dielectric constant values of the nanocomposites increase with increase of the CeO₂ content up to 30 wt% and decrease with further increase in the CeO₂ content. Besides this, there is good correlation between the dielectric constant values and LPG response. It is interesting to observe that the POA–CeO₂-30 nanocomposite exhibiting the maximum LPG response at room temperature has the maximum dielectric constant value. Thus, the LPG response of the sensors based on nanocomposites is supported and facilitated by the formation of p–n heterojunctions.

In addition to the gas response, the response and recovery properties are also important for evaluating the performance of gas sensors. The time taken by a sensor to achieve 90% of its maximum response is defined as response time in the case of adsorption or the recovery time in the case of desorption. The dynamic response–recovery characteristics of the sensors based on POA and POA–CeO₂-30 nanocomposite upon exposure to 3.4% LPG at room temperature are shown in Fig. 4(c). As can be seen from Fig. 4(c), the POA responded rapidly after introduction of LPG, however it does not recover, even after exposure to air for more than 5 minutes. The response time for the pure POA is about 3–4 s. On the other hand, the nanocomposite also responded very rapidly when exposed to the LPG and did not recover totally when it was exposed to air. The response and recovery times for the POA–CeO₂-30 nanocomposite are about ∼3–4 s and ∼20 s, respectively, which implies that both the response and recovery are rapid, which makes the present sensor suitable for the real time detection of LPG at room temperature. These values are lower than for the sensors based on polyaniline–ZnO,⁷ POA–SnO₂,³⁴ polythiophene–SnO₂,³⁵ polypyrrole–ZnO,³⁶ poly(o-chloroaniline)–ZnO,³⁷ core shell CeO₂ nanoparticle@cross-linked PANI nanohybrids⁴⁰ reported earlier. It is observed that the base line resistances of both the sensors are not recovered fully simply by exposing them to air and therefore the dynamic response–recovery characteristics [cf. Fig. 4(c)] are termed as irreversible. For irreversible gas-sensing, the surface reaction can be explained by the Langmuir–Hinshelwood mechanism and the response transient S(t) can be expressed by eqn (5)^54–56

S(t) = S_max(1 − e^−kC_at) (5)

where S_max is the maximum gas response when the sensor is saturated; k is the surface rate constant and C_a is the initial gas concentration. Solving eqn (5) for the response time (τ), the expression for τ can be obtained as

The response transients of POA and POA–CeO₂-30 for 3.4% LPG were fitted nonlinearly according to eqn (5) and the fitted curves are consistent with the experimental results in Fig. 4(c). The estimated values of the response times for both sensors are found to be ∼3–4 s. Interestingly, the POA sensor (Fig. S2(b), ESI†) recovers fully at 100 °C and its recovery time is found to be ∼40 s. The response and recovery times of the POA–CeO₂-30 nanocomposite sensor at 100 °C are found to be ∼3–4 s and 17–18 s, respectively. The recovery of the nanocomposite sensor is faster than that of the POA sensor (40 s), which is attributed to a faster rate of desorption at 100 °C.

From an application point of view, the reproducibility is the most important gas-sensing characteristic. It is the ability of the sensor to sequentially respond to the target gas without a visible change in the response. To examine the reproducibility of both sensors, the same sensors were exposed to 3.4% LPG at room temperature for 3 cycles. As can be seen from Fig. 4(d), the POA responded rapidly to LPG but it failed to return to the initial value after removal of the gas. Thus, the POA shows poor reversibility to LPG at room temperature. The sensing behaviour of the POA–CeO₂-30 based sensor is quite reproducible and stable. Furthermore, there is no clear change in the responses during the measurement of the three sensing cycles, exhibiting excellent stability and reproducibility of the sensor. The reproducible sensing behaviour indicates a weaker interaction between the LPG and sensing layer rather than any chemical reaction. Also, at 100 °C, the sensing behavior of both sensors is quite reproducible and the base line remains stable (Fig. S2(c), ESI†).

To further confirm the reliability for LPG detection, the performances of the sensors based on POA and POA–CeO₂-30 toward different LPG concentrations (0.14% to 4%) were investigated at room temperature. The experimental relationships between the responses and LPG concentration at room temperature are depicted in Fig. 5(a). The y-error bars represent the standard deviation at a fixed concentration and they are in the range of 0.01–0.06 and 0.02–0.13 for POA and POA–CeO₂-30, respectively, indicating the low experimental variability. It is seen that the responses of both sensors are highly dependent on the LPG concentration in the measurement range. Obviously, the nanocomposite sensor exhibits the higher response to LPG at different concentrations compared with that based on POA in the entire detecting range. As can be seen from Fig. 5(a), the responses of both sensors increase linearly with the increasing LPG concentration from 0.14% to 4%, however, the response rate of POA–CeO₂-30 nanocomposite is higher. The linear relationship between the gas response and LPG concentration can be expressed as the following eqn (7) and (8) for POA and POA–CeO₂-30 nanocomposite, respectively.

y = 0.018 + 0.19x, R² = 0.9919 (7)

y = 0.065 + 0.61x, R² = 0.9974 (8)

where x, y and R² represent the LPG concentration, gas response and correlation coefficient, respectively. The broken lines in the Fig. 5(a) indicate the linear fit to the experimental data, illustrating the good quality of the fit. Thus, the formation of the composite of POA with CeO₂ nanoparticles improves its response and it can be reliably used to monitor the concentration of LPG over this range.

Fig. 5 (a) Relationship between response and LPG concentration of pure POA and POA–CeO₂-30 nanocomposite based sensors. Dots represent the experimental data points, and the discontinuous lines represent a linear fit to the experimental data; (b) bar chart showing the gas responses of pure POA and POA–CeO₂-30 nanocomposite for different gases at room temperature. The gas concentration in all cases was 3.4%; (c) stability of the POA–CeO₂-30 nanocomposite sensor and (d) response of the POA–CeO₂-30 nanocomposite sensor towards relative humidity. Error bars indicates the standard deviation in the measurement.

Detection limit is another important factor used for the evaluation of the sensing-performance of the sensors. It is defined as the lowest concentration of the analyte at which the response of the sensor under the given conditions is differentiated from the background level.^50,52,57 When this criterion is applied in the present work, the LPG response was set to S > 0.088. According to eqn (7) and (8), the detection limits are estimated to be approximately 0.13% and 0.09% for the sensors based on pure POA and POA–CeO₂-30 nanocomposite, respectively. Thus, the detection limit for POA–CeO₂-30 nanocomposite (∼0.09%) is lower than the PEL for LPG. The similar variation of the response with the LPG concentration can also be obtained at the operating temperature of 100 °C (Fig. S2(d), ESI†), implying the excellent sensing performance at higher temperature. This result demonstrates that the formation of a composite of POA with CeO₂ nanoparticles lowers the detection limit of POA and is capable of detecting a concentration of LPG lower than PEL limit.

Selectivity in gas-sensing is another important parameter of the sensor, which relies mainly on the specific interactions between the sensing layer and the target gas molecules. It is defined as the ability of a sensor to respond to a target gas in the presence of other interfering gases. The LPG sensing selectivity of POA and POA–CeO₂-30 based sensors was examined towards various gases such as hydrogen (H₂), carbon monoxide (CO), carbon dioxide (CO₂) and ethanol (C₂H₅OH) at 3.4% and room temperature. The responses to each gas were calculated by eqn (1). Fig. 5(b) depicts the histogram of the responses of two sensors based on POA and POA–CeO₂-30 nanocomposite toward 3.4% of LPG, H₂, CO, CO₂ and ethanol at room temperature. The y-error bars represent the standard deviation at a fixed target gas and they are in the range of 0.06–0.07 and 0.07–0.10 for POA and POA–CeO₂-30, respectively, indicating the low experimental variability. The POA based sensor shows almost nearly equal response values to all the test gases. Besides, by contrast, the sensor based on the POA–CeO₂-30 shows enhanced responses to all the test gases compared with its pure POA counterpart. However, it exhibits a high response of 2.06% to 3.4% of LPG compared to other test gases, indicating that the POA–CeO₂-30 nanocomposite has a better ability to discriminate LPG. In order to quantify the selectivity to LPG, the selectivity coefficient (K) was calculated further according to definition^34,58

where S_LPG and S_B are the responses of sensors to LPG and B gas, respectively. The higher value of K implies the more selective detection of LPG in the presence of other gases. The selectivity coefficients for the POA based sensor are found to be 1.46 to H₂, 1.45 to CO, 1.49 to CO₂, and 1.63 to ethanol, indicating the lack of selectivity to the LPG sensing at room temperature. Even at the operating temperature of 100 °C, the POA sensor shows a lack of LPG sensing selectivity (Fig. S3, ESI†). On the contrary, the values of K for the POA–CeO₂-30 based sensor are 4.90, 4.29, 3.68 and 3.68 towards H₂, ethanol, CO and CO₂, respectively. The result suggests that the POA–CeO₂-30 sensor exhibits LPG sensing selectivity at room temperature. Notably, in contrast to pure POA, the POA–CeO₂-30 sensor still shows an excellent LPG sensing selectivity (Fig. S3, ESI†) even at the operating temperature of 100 °C. Therefore, the POA–CeO₂-30 nanocomposite sensor is suitable for selective detection of LPG at room temperature. The different adsorption abilities and interactions of target gases on the POA–CeO₂-30 nanocomposite are different at a given operating temperature, which may result in the superior LPG sensing selectivity. Thus, the formation of a composite of POA with CeO₂ nanoparticles is not only effective in enhancing the gas response but also in making it selective for the detection of LPG.

The POA–CeO₂-30 sensor was also tested for its stability in ambient conditions. To investigate its stability, the sensor was kept in atmospheric conditions for 30 days and its sensing-performance towards 3.4% LPG at room temperature was measured under the identical experimental conditions. Fig. 5(c) shows the comparison of the sensing-performance towards 3.4% LPG at room temperature before and after 30 days of storage in air. As seen from Fig. 5(c), the sensor still responded well towards 3.4% LPG with slight degradation after 30 days and the standard deviation in the LPG response values was observed to be ∼0.18, which directly confirms the excellent stability of the sensor. However, when the sensors were stored in air for a relatively long time, their performances decrease dramatically, which can be attributed to the dedoping (undoping) of conducting polymers. Generally, conducting polymers are used mainly in resistive-type sensors and, therefore, the change in film conductivity causes both the drift of sensor baseline and the change of sensor response. It is well-known that the stability of conductivity in air strongly depends on the dopants.^59,60 Therefore, a possible mechanism to increase the stability of the conducting polymer (and their composites) based sensors is to dope the conducting polymers with big anions such as H₂SO₄, p-toluene sulfonic acid (p-TSA) and camphorsulfonic acid (CSA). Another approach to improve the stability of the conducting polymer based sensors is the stabilization of the polymers, which can be achieved by the formation of composites of two conducting polymers and copolymers by oxidative chemical polymerization. As mentioned earlier, Wang et al. synthesized core–shell CeO₂ nanoparticles@cross-linked PANI nanohybrids by a facile in situ polymerization method using phytic acid, which worked as gelator and dopant simultaneously. The gelator phytic acid, due to its nonvolatile nature, worked as a cross-linker to weaken the impact of the environment by increasing the mechanical strength and chemical stability of the sensor, which combined excellent sensitivity with superior long-term stability.

Furthermore, no significant variation in the response and recovery abilities was observed. It was also demonstrated that humidity has a great effect on the performance of conducting polymer⁶⁰ and CeO₂ (ref. 42) based sensors. Therefore, the study of the influence of humidity on the fabricated sensor based on POA–CeO₂ nanocomposite is important for its real-time practical applications. In order to investigate the influence of humidity on the POA–CeO₂-30 nanocomposite sensor, the experiments were also performed to measure the resistance of the pellet as a function of relative humidity by using a simple two probe configuration with an electrometer (6517B Electrometer, Keithley) controlled by the test software supplied by Biotronic systems, Mumbai, India. In these experiments, the different relative humidity (RH) levels 11 ± 0.30, 33 ± 0.14, 43 ± 0.20, 75 ± 0.15, 85 ± 60.24, and 97 ± 0.16% RH were generated in a simple gas-sensing setup fabricated in our laboratory. The variation of the resistance of the POA–CeO₂-30 nanocomposite with the relative humidity is shown in Fig. 5(d). The y-error bars represent the standard deviation at a fixed RH and they are in the range of 0.018–0.020, indicating the low experimental variability. It is seen that the resistance of the nanocomposite did not show a significant variation when the humidity was increased from 11% to 97% RH. This may be attributed to the incorporation of CeO₂ nanoparticles in the POA matrix, thereby reducing the bare POA sites for the adsorption of water molecules.⁶¹ Thus, the results further confirmed that the present sensor can have a practical application on the basis of its excellent LPG sensing selectivity and good stability.

It is well known that the gas-sensing response is the reflection of the kinetics of the gas adsorption and desorption processes occurring on the surface of the sensor.^62,63 As a result, the response and recovery times depend on the rate of adsorption and desorption respectively. The gas adsorption and desorption processes can be studied by transforming the dynamic response–recovery characteristics as shown in Fig. 4(c) using a derivative method proposed by Zhang et al.⁶⁴ In this method, the first time derivative of the response can be used to represent the adsorption–desorption kinetics during the sensing process and it is plotted as a function of the response (S). Physically, S reflects the concentration of adsorbed gaseous species, while represents the rate of change of concentration of gaseous species on the sensor surface under certain measuring conditions. A typical gas adsorption–desorption curve is depicted in Fig. 6(a), where the upper and lower parts of the curve represent the adsorption and desorption cycles, respectively occurring during the gas-sensing process. Part A of the adsorption cycle represents adsorption and diffusion of gaseous species on the sensor surface during the initial sensing stage and its slope gives the speed of adsorption. As the gas-sensing process persists, the adsorption of gaseous species slows down gradually and adsorption cycle becomes relatively horizontal following the linearly upward portion of part A. Subsequently, the adsorption and desorption processes reach to a equilibrium and the attains a maximum value (h₁) at a certain value of gas response S₁. Once the equilibrium is reached, the desorption of the products produced in the gas-sensing process starts playing a dominating role and the gas response increases to the maximum value S₀, which represents part B of the adsorption cycle. The length of part B of the adsorption process is S₂ and the sum S₁ + S₂ = S₀ represents the length of the adsorption process. After the removal of the gas, the desorption of the gas dominates the gas-sensing process and the attains a minimum value (h₂) at a certain value of gas response S₁ and the response of the gas sensor returns to the baseline within the length S₂, leading to the recovery of the sensor. The sum S′₁ + S′₂ is the length of the desorption process. Thus, the important physical parameters to be extracted from the adsorption–desorption curve [Fig. 6(a)] are maximum gas response or length of adsorption process (S₀), the height of adsorption cycle (h₂), the location of h₁(S₁), the height of desorption cycle (h₂) and the location of (h₂)(S′₁). Using the transformation method as described above, the dynamic response–recovery characteristics of the sensor based on POA in Fig. 4(c) were transformed to the curve having parameters S₀ = 0.81, h₁ = 0.34, S₁ = 0.41, h₂ = 0.18 and S′₁ = 0.21, which is shown in Fig. 6(b). When the sensor based on POA is exposed to LPG at room temperature, the LPG molecules get adsorbed and diffuse rapidly, as indicated by the sharp rise of the adsorption curve. This is attributed to the porous structure of pure POA, which permits large penetration of LPG molecules into the entire volume of the sensing layer. As the gas-sensing process goes on, the adsorption of the LPG gas slows down gradually, thereby exhibiting a horizontal portion. When the gas adsorption and desorption reach equilibrium, the LPG response increases to the maximum value, S₀ = 0.81. Thus, the length of the adsorption process is ∼0.81. After the removal of the LPG, the desorption of gas molecules dominates the gas-sensing process, resulting in the decrease of the response which leads to the recovery of the sensor. In the initial stage, the rate of desorption is fairly high, however, it is lower than that of adsorption as seen by a smaller value. Moreover, the desorption cycle remains incomplete as indicated by negative values, suggesting that desorption by diffusion of product gases not only takes a longer time, but also remains incomplete. This may be due to the fact that, LPG molecules being heavier in nature, the reaction products are not leaving immediately from POA. This illustrates why the POA based sensor does not recover even after exposure to air for more than 5 minutes. Due to a faster rate of desorption at 100 °C, the desorption cycle (Fig. S4(a), ESI†) gets completed, thereby indicating that the reaction products diffused out easily and the POA based sensor recovers fully with recovery time of 40 s.

Fig. 6 (a) A typical gas adsorption–desorption curve of the gas sensor; gas adsorption–desorption curves of the sensors based on (b) pure POA and (c) POA–CeO₂-30 nanocomposite toward 3.4% LPG at room temperature.

The dynamic response–recovery characteristics of the sensor based on POA–CeO₂-30 nanocomposite in Fig. 4(c) were also transformed to the curve having parameters S₀ = 2.04, h₁ = 0.78, S₁ = 0.95, h₂ = 0.22 and S′₁ = 0.21, which is shown in Fig. 6(c). When the sensor based on POA–CeO₂-30 is exposed to LPG at room temperature, the porosity of POA provides a path for LPG molecules to reach the active sites of CeO₂ nanoparticles on the backbone of POA, which is indicated by the initial steep rise of the adsorption curve. As compared with the kinetic process of gas adsorption–desorption of POA, the POA–CeO₂-30 nanocomposite exhibits an inflected gas adsorption–desorption curve having parameters with higher values. The observed inflection in the gas adsorption–desorption curve is attributed to the existence of p–n heterojunctions in the nanocomposite coupled with an increase in the surface for the interaction with LPG molecules. In contrast to POA, the desorption cycle is completed after the removal of the LPG. Moreover, the rate of desorption is lower than that of adsorption, indicating that the desorption and diffusion of product gases take a longer time than the adsorption and diffusion of LPG gas. This illustrates why the recovery time of the POA–CeO₂-30 based sensor is longer than the response time. At 100 °C, the POA–CeO₂-30 nanocomposite sensor reveals inflected gas adsorption–desorption curves (Fig. S4(b), ESI†) having parameters with higher values compared with the room temperature curves (Fig. 6(c)) and the curves of the POA based sensor at 100 °C (Fig. S4(a), ESI†). Careful comparison with Fig. 6(c), suggests that the rates of adsorption and desorption are accelerated by almost 40 and 20 times, respectively.

Collectively, the present sensor shows a higher LPG response (∼2.06% to 3.4% LPG), room temperature operation, excellent reproducibility (3 cycles with highest response), rapid response (∼3–4 s), fast recovery (∼20 s), good sensing selectivity, stability (stable up to 30 days) and lower detection limit (∼0.09% LPG < PEL). Overall, the LPG sensing characteristics of the present sensor in this study are better than other competing conducting polymer–metal oxide nanocomposite based LPG sensors reported in the literature,^7,34–37,40 which are compared and summarized in Table 1. But the sensing conditions such as operating temperature, LPG concentration, sensor configuration, synthesis method and properties of the nanocomposites were different. Hence, comprehensively considering the LPG sensing characteristics from Table 1, it is obvious that the POA–CeO₂-30 nanocomposite sensor possesses superior properties than those reported in the literature. In short, the sensor based on POA–CeO₂ nanocomposites can overcome the shortcomings of long response time, slow recovery, lack of selectivity of POA and higher operating temperature of CeO₂, thus presenting a very promising sensing material for the fabrication of LPG sensors operating at room temperature.

Table 1 Summary of LPG sensing performances of various conducting polymer–metal oxide nanocomposite based sensors^a

No.	Composite	LPG response	Operating temperature	Response/recovery times (s)	Selectivity	Detection limit	Reproducibility	Stability	Reference
a RT: room temperature, NS: not studied.
1	Poly(o-anisidine)–SnO2	23.47% to 3.4%	100 °C	5–6 s/33 s	Remarkable selectivity	NS	3 cycles	NS	34
2	Polythiophene–SnO2	56.2% to 2.5%	RT	94 s/466 s	NS	NS	NS	NS	35
3	Polypyrrole–ZnO	32.5% to 1400 ppm	RT	3.5–4 min/40 min	NS	NS	3 cycles	NS	36
4	Poly(o-chloroaniline)–ZnO	93% to 400 ppm	RT	56.76 s/37.59 s	Remarkable selectivity	NS	5 cycles	60 days	37
5	Polyaniline–ZnO	8.73% to 1000 ppm	36 °C	100 s/185 s	NS	NS	NS	NS	7
6	Poly(o-anisidine)–CeO2	2.06% to 3.4%	RT	3–4 s/20 s	Remarkable selectivity	0.09%	3 cycles	30 days	Present work

---

3.3 LPG sensing mechanism

According to the experimental results presented above, the POA–CeO₂ nanocomposite sensor exhibits enhanced response and good sensing selectivity to LPG at room temperature, compared with the POA sensor, which is attributed to the effect of the p–n heterojunctions formed between the interface of the POA and CeO₂ nanoparticles. Fig. 7(a) and (b) shows the energy band structure of POA–CeO₂ heterojunction before and after the contact. The experimental measured values of band gap energies, the positions of the LUMO, HOMO, conduction band and valence band are used to construct the energy band diagram. The energies of the band gap, LUMO and HOMO of POA are 3.85, 0.4 and 4.25, respectively.⁶⁵ The band gap energy and the electron affinity of CeO₂ are 3.8 and 4.69 eV, respectively.⁶⁶ The E_F, E_C and E_V denote the Fermi level and the positions of the conduction and valence bands, respectively. After the polymerization of o-anisidine on the surface of CeO₂ nanoparticles, the p–n heterojunctions will be formed between the p-POA and n-CeO₂ nanoparticles. The electrons transfer from CeO₂ to POA while the holes transfer from POA to CeO₂ until both Fermi levels become equal, which is a Fermi level-mediated charge transfer. Finally, an electron depletion layer is formed at the interface between POA and CeO₂, leading to a heterojunction barrier due to band bending, as shown in Fig. 7(b).

Fig. 7 Energy band diagrams for (a) POA and CeO₂ nanoparticles, (b) POA–CeO₂ nanocomposite and (c) schematic model for the POA–CeO₂ nanocomposite sensor when exposed to air and LPG at room temperature.

Fig. 7(c) presents a schematic model for the POA–CeO₂ nanocomposite sensor when exposed to air and LPG at room temperature. As mentioned earlier, the CeO₂ nanoparticles are enwrapped and interconnected with POA in the nanocomposite and therefore, the active sites on the POA backbone are largely accessible to air and LPG. When the nanocomposite sensor is exposed to air, the surface of POA (p-type semiconductor) is readily covered with chemisorbed oxygen, even at low oxygen partial pressure and room temperature. Thus, at room temperature, in the absence of LPG, oxygen gets adsorbed on the surface of POA and it extracts electrons from its LUMO, which can be explained by the following reactions^67–69

O₂ (gas) ↔ O₂ (adsorbed) (10)

O₂ (adsorbed) + e⁻ → O₂⁻ (chemisorbed) (11)

Here, the oxygen molecules are continuously supplied from the dry air and are adsorbed on the surface of the nanocomposite. The electron transfer from the LUMO to the chemisorbed oxygen results in an increase in the hole density. As a consequence, a decrease in the resistance of the nanocomposite is observed and the equilibration of the chemisorption process results in stabilization of the surface resistance. The stabilized resistance in this state is known as the resistance of the sensor in the presence of air (R_a). Any process that disturbs this equilibrium gives rise to changes in the resistance of the nanocomposite sensor. The LPG sensing process is based on the changes in the resistance of the nanocomposite sensor which is controlled by the LPG species and the amount of the chemisorbed oxygen on the surface.

It is well known that LPG consists of CH₄, C₃H₈ and C₄H₁₀ etc. and in these molecules the reducing hydrogen species are bound to carbon, therefore LPG dissociates less easily into the reactive reducing components on the nanocomposite surface. When the nanocomposite sensor is exposed to LPG, it can be partially oxidized through the reaction with the chemisorbed O₂⁻ on the surface of the sensor. The overall reaction can be written as follows^8,13,70,71

C_nH_2n+2 + O₂⁻ (chemisorbed) → H₂O (adsorbed) + C_nH_2nO + e⁻ (12)

where C_nH_2n+2 includes CH₄, C₃H₈ and C₄H₁₀; C_nH_2nO is the partially oxidized intermediates.⁶⁹ The free electrons released via the interaction between the LPG molecules and the chemisorbed O₂⁻ neutralize the holes, which are the majority carriers in p-type POA. This compensation results in a decrease in the hole carriers in POA and results in more negative acceptor ions in the p-side and hence a wider space charge layer. As a result, the carrier concentration of the heterojunction decreases and, consequently, the potential barrier height of the heterojunction increases and an increase in the resistance of the nanocomposite is observed.

When the gas-sensing chamber is exposed to air for recovery, the oxygen molecules present in air will adsorb on the nanocomposite surface and capture the electrons through the processes indicated in eqn (10) and (12), resulting in the decrease in the resistance of the nanocomposite sensor. It can be said that in the POA–CeO₂ nanocomposite, POA identifies LPG (receptor function) and both POA and CeO₂ nanoparticles provide a conducting path (transducer function) and their nanocomposite permits selective detection of LPG at room temperature.

4. Conclusions

The application of POA–CeO₂ nanocomposite for the detection of LPG at room temperature is successfully demonstrated. The POA–CeO₂ nanocomposite based sensor showed enhanced LPG sensing characteristics compared with those of individual POA and CeO₂ nanoparticles. The enhanced LPG response (∼2.06% to 3.4% LPG), excellent reproducibility, rapid response (∼3–4 s), fast recovery (∼20 s), good sensing selectivity, stability (stable up to 30 days) and detection limit lower than PEL have been achieved by the POA–CeO₂ nanocomposite sensor. The POA–CeO₂ nanocomposite sensor is, thus, cost-effective, room temperature operable, easy to fabricate, selective and it possesses all the essential characteristics of a reliable sensor.

Acknowledgements

The authors acknowledge financial support from University Grants Commission (UGC), New Delhi (Project No. F. 530/2/DRS/2010-SAP-DRS, Phase-II). Sharada A. Thakur is thankful to UGC, New Delhi for awarding the Rajiv Gandhi National Fellowship.

References

1. N. Gaikwad, S. Bhanoth, P. V. More, G. H. Jain and P. K. Khanna, Chemically designed Pt/PPy nano-composite for effective LPG gas sensor, Nanoscale, 2014, 6, 2746–2751.
2. H. L. Maclean and L. B. Lave, Evaluating automobile fuel/propulsion system technologies, Prog. Energy Combust. Sci., 2003, 29, 1–69.
3. C. C. Chang, J. G. Lo and J. L. Wang, Assessment of reducing ozone forming potential for vehicles using liquefied petroleum gas as an alternative fuel, Atmos. Environ., 2001, 35, 6201–6211.
4. W. Z. Qian, H. Yu, F. Wei, Q. F. Zhang and Z. W. Wang, Synthesis of carbon nanotubes from liquefied petroleum gas containing sulfur, Carbon, 2002, 40, 2968–2970.
5. A. R. Phani, X-ray photoelectron spectroscopy studies on Pd doped SnO₂ liquid petroleum gas sensor, Appl. Phys. Lett., 1997, 71, 2358–2360.
6. M. Batzill and U. Diebold, Surface studies of gas sensing metal oxides, Phys. Chem. Chem. Phys., 2007, 9, 2307–2318.
7. P. T. Patil, R. S. Anwane and S. B. Kondawar, Development of electrospun polyaniline/ZnO composite nanofibers for LPG sensing, Procedia Mater. Sci., 2015, 10, 195–204.
8. Y. Fu, Y. Nie, Y. Zhao, P. Wang, L. Xing, Y. Zhang and X. Xue, Detecting liquefied petroleum gas (LPG) at room temperature using ZnSnO₃/ZnO nanowire piezo-nanogenerator as self-powered gas sensor, ACS Appl. Mater. Interfaces, 2015, 7, 10482–10490.
9. D. Patil, V. Patil and P. Patil, Highly sensitive and selective LPG sensor based on α-Fe₂O₃ nanorods, Sens. Actuators, B, 2011, 152, 299–306.
10. R. B. Waghulade, R. Pasricha and P. P. Patil, Synthesis and LPG sensing properties of nano-sized cadmium oxide, Talanta, 2007, 72, 594–599.
11. P. P. Sahay and R. K. Nath, Al-doped zinc oxide thin films for liquid petroleum gas (LPG) sensors, Sens. Actuators, B, 2008, 133, 222–227.
12. G. N. Chaudhari, D. R. Bambole and A. B. Bodade, Structural and gas sensing behavior of nanocrystalline BaTiO₃ based liquid petroleum gas sensors, Vacuum, 2006, 81, 251–256.
13. S. Singh, B. C. Yadav, V. D. Gupta and P. K. Dwivedi, Investigation on effects of surface morphologies on response of LPG sensor based on nanostructured copper ferrite system, Mater. Res. Bull., 2012, 47, 3538–3547.
14. R. B. Birajadar, D. Upadhye, S. Mahajan, J. C. Vyas and R. Sharma, J. Mater. Sci.: Mater. Electron., 2015, 26, 5065–5070.
15. Y. Liu, J. Parisi, X. Sun and Y. Lei, Study of room temperature LPG sensing behaviour of polyaniline thin films synthesized by cost effective oxidative polymerization technique, J. Mater. Chem. A, 2014, 2, 9919–9943.
16. E. Comini, Metal oxide nano-crystals for gas sensing, Anal. Chim. Acta, 2006, 568, 28–40.
17. I. Fratoddia, I. Vendittia, C. Camettib and M. V. Russoa, Chemiresistive polyaniline-based gas sensors: A mini review, Sens. Actuators, B, 2015, 220, 534–548.
18. H. Bai and G. Shi, Gas sensors based on conducting polymers, Sensors, 2007, 7, 267–307.
19. T. Ahuja and D. Kumar, Recent progress in the development of nano-structured conducting polymers/nanocomposites for sensor applications, Sens. Actuators, B, 2009, 136, 275–286.
20. K. Ramanathan, M. A. Bangar, M. Yun, W. Chen, A. Mulchandani and N. V. Myung, Individually addressable conducting polymer nanowires array, Nano Lett., 2004, 4, 1237–1239.
21. J. Wang, S. Chan, R. R. Carlson, Y. Luo, G. Ge, R. S. Ries, J. R. Heath and H. R. Tseng, Electrochemically fabricated polyaniline nanoframework electrode junctions that function as resistive sensors, Nano Lett., 2004, 4, 1693–1697.
22. S. Bhadra, D. Khastgir, N. K. Singha and J. H. Lee, Progress in preparation, processing and applications of polyaniline, Prog. Polym. Sci., 2009, 34, 783–810.
23. A. Z. Sadeka, W. Wlodarskia, K. Kalantar-Zadeha, C. Bakerb and R. B. Kaner, Doped and dedoped polyaniline nanofiber based conductometric hydrogen gas sensors, Sens. Actuators, A, 2007, 139, 53–57.
24. V. Dixit, S. C. K. Misra and B. S. Sharma, Carbon monoxide sensitivity of vacuum deposited polyaniline semiconducting thin films, Sens. Actuators, B, 2005, 104, 90–93.
25. Z. Jin, Y. Su and Y. Duan, Development of a polyaniline-based optical ammonia sensor, Sens. Actuators, B, 2001, 72, 75–79.
26. E. Kriván, C. Visy, R. Dobay, G. Harsányi and O. Berkesi, Irregular response of the polypyrrole films to H₂S, Electroanalysis, 2000, 12, 1195–1200.
27. X. B. Yan, Z. J. Han, Y. Yang and B. K. Tay, NO₂ gas sensing with polyaniline nanofibers synthesized by a facile aqueous/organic interfacial polymerization, Sens. Actuators, B, 2007, 123, 107–113.
28. A. Kaushik, R. Kumar, S. K. Arya, M. Nair, B. D. Malhotra and S. Bhansali, Organic–inorganic hybrid nanocomposite-based gas sensors for environmental monitoring, Chem. Rev., 2015, 115, 4571–4606.
29. Rajesh, T. Ahuja and D. Kumar, Sens. Actuators, B, 2009, 136, 275–286.
30. H. Tai, Y. Jiang, G. Xie, J. Yu and X. Chen, Recent progress in the development of nano-structured conducting polymers/nanocomposites for sensor applications, Sens. Actuators, B, 2007, 125, 644–650.
31. H. Tai, Y. Jiang, G. Xie, J. Yu, X. Chen and Z. Ying, Influence of polymerization temperature on NH₃ response of PANI/TiO₂ thin film gas sensor, Sens. Actuators, B, 2008, 129, 319–326.
32. L. Geng, Y. Zhao, X. Huang, S. Wang, S. Zhang and S. Wu, Characterization and gas sensitivity study of polyaniline/SnO₂ hybrid material prepared by hydrothermal route, Sens. Actuators, B, 2007, 120, 568–572.
33. L. Geng, X. Huang, Y. Zhao, P. Li, S. Wang, S. Zhang and S. Wu, H₂S sensitivity study of polypyrrole/WO₃ materials, Solid-State Electron., 2006, 50, 723–726.
34. D. Patil, K. Kolhe, H. S. Potdar and P. Patil, Investigation of poly(o-anisidine)–SnO₂ nanocomposites for fabrication of low température operative liquefied petroleum gas sensor, J. Appl. Phys., 2011, 110, 124501–124508.
35. S. S. Barkade, D. V. Pinjarib, U. T. Nakate, A. K. Singh, P. R. Gogate, J. B. Naik, S. H. Sonawane and A. B. Pandit, Ultrasound assisted synthesis of polythiophene/SnO₂ hybrid nanolatex particles for LPG sensing, Chem. Eng. Process., 2013, 74, 115–123.
36. S. S. Barkade, D. V. Pinjari, A. K. Singh, P. R. Gogate, J. B. Naik, S. H. Sonawane, M. Ashokkumar and A. B. Pandit, Ultrasound assisted miniemulsion polymerization for preparation of polypyrrole–zinc oxide (PPy/ZnO) functional Latex for liquefied petroleum gas sensing, Ind. Eng. Chem. Res., 2013, 52, 7704–7712.
37. S. Khasim and O. A. Al-Hartomy, Fabrication and gas sensitivity in heterostructures of ortho-chloropolyaniline–ZnO nanocomposites, RSC Adv., 2014, 4, 39844–39852.
38. C. R. Michela and A. H. Martínez-Preciado, CO sensor based on thick films of 3D hierarchical CeO₂ architectures, Sens. Actuators, B, 2014, 197, 177–184.
39. R. Benea, I. V. Perczel, F. ReÂti, F. A. Meyer, M. Fleisher and H. Meixner, Chemical reactions in the detection of acetone and NO by a CeO₂ thin film, Sens. Actuators, B, 2000, 71, 36–41.
40. L. Wang, H. Huang, S. Xiao, D. Cai, Y. Liu, B. Liu, D. Wang, C. Wang, H. Li, Y. Wang, Q. Li and T. Wang, Enhanced sensitivity and stability of room-temperature NH₃ sensors using core–shell CeO₂ nanoparticles@cross-linked PANI with p–n Heterojunctions, ACS Appl. Mater. Interfaces, 2014, 6, 14131–14140.
41. B. S. Khened, T. Machappa, M. V. N. Ambika Prasad and M. Sasikala, Studies on DC conductivity and LPG sensing behavior of PANI/CeO₂ composites, Int. J. Sci. Res., 2013, 1, 180–184.
42. S. Thakur and P. Patil, Rapid synthesis of cerium oxide nanoparticles with superior humidity-sensing performance, Sens. Actuators, B, 2014, 194, 260–268.
43. J. Liu, F. Wang and R. Dewil, CeO₂ nanocrystalline-supported palladium chloride: an effective catalyst for selective oxidation of alcohols by oxygen, Catal. Lett., 2009, 130, 448–454.
44. L. Yue and X. M. Zhang, Structural characterization and photocatalytic behaviours of doped CeO₂ nanoparticles, J. Alloys Compd., 2009, 475, 702–705.
45. D. Patil, P. Patil, Y. K. Seo and Y. K. Hwang, Poly(o-anisidine)–tin oxide nanocomposite: synthesis, characterization and application to humidity sensing, Sens. Actuators, B, 2010, 148, 41–48.
46. D. Patil, Y. K. Seo, Y. K. Hwang, J. S. Chang and P. Patil, Humidity sensing properties of poly(o-anisidine)/WO₃ composites, Sens. Actuators, B, 2008, 128, 374–382.
47. E. M. Genies, A. Boyle, M. Lapkowski and C. Tsintavis, Polyaniline: a historical survey, Synth. Met., 1990, 36, 139–182.
48. W. Feng, E. Sun, A. Fujii, H. Wu, K. Niihara and K. Yoshino, Synthesis and characterization of photoconducting polyaniline–TiO₂ nanocomposite, Bull. Chem. Soc. Jpn., 2000, 73, 2627–2633.
49. E. Kumar, P. Selvarajan and D. Muthuraj, Preparation and characterization of polyaniline/cerium dioxide (CeO₂) nanocomposite via in situ polymerization, J. Mater. Sci., 2012, 47, 7148–7156.
50. C. Murugan, E. Subramanian and D. P. Padiyan, p–n heterojunction formation in polyaniline–SnO₂ organic–inorganic hybrid composite materials leading to enhancement in sensor functionality toward benzene and toluene vapors at room temperature, Synth. Met., 2014, 192, 106–112.
51. J. Zhang, S. Wang, M. Xu, Y. Wang, H. Xia, S. Zhang, X. Guo and S. Wu, Polypyrrole coated SnO₂ hollow spheres and their application for ammonia sensor, J. Phys. Chem. C, 2009, 113, 1662–1665.
52. C. Murugan, E. Subramanian and D. P. Padiyan, Enhanced sensor functionality of in situ synthesized polyaniline–SnO₂ hybrids toward benzene and toluene vapors, Sens. Actuators, B, 2014, 205, 74–81.
53. T. J. Savenije, E. Moons, G. K. Boschloo, A. Goossens and T. J. Schaafsma, Photogeneration and transport of charge carriers in a porphyrin p/n heterojunction, Phys. Rev. B: Condens. Matter Mater. Phys., 1997, 55, 9685–9692.
54. S. Bai, Y. Tian, M. Cui, J. Sun, Y. Tian, R. Luo, A. Chen and D. Li, Polyaniline@SnO₂ heterojunction loading on flexible PET thin film for detection of NH₃ at room temperature, Sens. Actuators, B, 2016, 226, 540–547.
55. H. B. Zeng, G. T. Duan, Y. Li, K. Yang, X. X. Xu and W. P. Cai, Blue luminescence of ZnO nanoparticles based on non-equilibrium processes: defect origins and emission controls, Adv. Funct. Mater., 2010, 20, 561–572.
56. C. Y. Lee and M. S. Strano, Understanding the dynamics of signal transduction for adsorption of gases and vapors on carbon nanotube sensors, Langmuir, 2005, 21, 5192–5196.
57. R. Pandeeswari and B. G. Jeyaprakash, Understanding the dynamics of signal transduction for adsorption of gases and vapors on carbon nanotube sensors, Sens. Actuators, B, 2014, 195, 206–214.
58. M. Siemons and U. Simon, Gas sensing properties of volume-doped CoTiO₃ synthesized via polyol method, Sens. Actuators, B, 2007, 126, 595–603.
59. H. L. Ricks-Laskoski and L. J. Buckley, Twenty-year aging study of electrically conductive polypyrrole films, Synth. Met., 2006, 156, 417–419.
60. L. Bay, N. Mogensen, S. Skaarup, P. Sommer-Larsen, M. Jørgensen and K. West, Polypyrrole doped with alkyl benzenesulfonates, Macromolecules, 2002, 35, 9345–9351.
61. V. Chaudhari and A. Kaur, Enhanced room temperature sulfur dioxide sensing behaviour of in situ polymerized polyaniline–tungsten oxide nanocomposite possessing honeycomb morphology, RSC Adv., 2015, 5, 73535–73544.
62. J. Liu, T. Luo, F. Meng, K. Qian, Y. Wan and J. Liu, Porous hierarchical In₂O₃ micro-/nanostructures: preparation, formation mechanism, and their application in gas sensors for noxious volatile organic compound detection, J. Phys. Chem. C, 2010, 114, 4887–4894.
63. S. Bai, K. Zhang, L. Wang, J. Sun, R. Luo, D. Li and A. Chen, Synthesis mechanism and gas-sensing application of nanosheet-assembled tungsten oxide microspheres, J. Mater. Chem. A, 2014, 2, 7927–7934.
64. S. Zhang, C. Xie, M. Hu, H. Li, Z. Bai and D. Zeng, An entire feature extraction method of metal oxide gas sensors, Sens. Actuators, B, 2008, 132, 81–89.
65. P. Routh, P. Mukherjee and A. K. Nandi, Enhanced optoelectronic properties of RNA–poly(o-methoxyaniline) hybrid containing monodispersed Au nanoparticles, J. Mater. Chem., 2010, 20, 7214–7221.
66. S. S. Warule, N. S. Chaudhari, B. B. Kale, K. R. Patil, P. M. Koinkar, M. A. More and R. Murakami, Organization of cubic CeO₂ nanoparticles on the edges of self assembled tapered ZnO nanorods via a template free one-pot synthesis: significant cathodoluminescence and field emission properties, J. Mater. Chem., 2012, 22, 8887–8895.
67. Y. Liu, Y. Jiao, Z. Zhang, F. Qu, A. Umar and X. Wu, Hierarchical SnO₂ nanostructures made of intermingled ultrathin nanosheets for environmental remediation, smart gas sensor, and supercapacitor applications, ACS Appl. Mater. Interfaces, 2014, 6, 2174–2184.
68. C. L. Hsu, Y. D. Gao, Y. S. Chen and T. J. Hsueh, Vertical p-type Cu-doped ZnO/n-type ZnO homojunction nanowire-based ultraviolet photodetector by the furnace system with hotwire assistance, ACS Appl. Mater. Interfaces, 2014, 6, 4277–4285.
69. L. L. Xing, C. H. Ma, Z. H. Chen, Y. J. Chen and X. Y. Xue, High gas sensing performance of one-step-synthesized Pd–ZnO nanoflowers due to surface reactions and modifications, Nanotechnology, 2011, 22, 215501–215508.
70. S. Singh, A. Singh, B. C. Yadav and P. K. Dwivedi, Fabrication of nanobeads structured perovskite type neodymium iron oxide film: Its structural, optical, electrical and LPG sensing investigations, Sens. Actuators, B, 2013, 177, 730–739.
71. S. Singh, V. D. Gupta, B. C. Yadav, P. Tandon and A. K. Singh, Structural analysis of nanostructured iron antimonate by experimental and quantum chemical simulation and its LPG sensing, Sens. Actuators, B, 2014, 195, 373–381.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra02955h

---