Polyvinyl alcohol–In₂O₃ nanocomposite films: synthesis, characterization and gas sensing properties

Abstract

Poly(vinyl alcohol) (PVA)–In₂O₃ (with 1 and 5 wt% In₂O₃ loading) nanocomposite films have been prepared by a solvent-casting technique. The In₂O₃ nanoparticles used in this work were prepared by nonhydrolytic alcoholysis ester elimination reaction of indium acetate in the presence of oleic acid and oleyl alcohol at 220 °C. X-Ray diffraction (XRD) patterns and transmission electron microscopy (TEM) studies indicate that the In₂O₃ nanocrystals obtained in this work are nearly monodisperse, highly crystalline with cubic bixbyite structure without the presence of any other impurity phase. The PVA–In₂O₃ nanocomposite films have been structurally characterized by XRD, Fourier transform infrared (FTIR) and Raman spectroscopy. The results confirm the incorporation of In₂O₃ nanocrystals in the PVA matrix and interactions between In₂O₃ nanocrystals and PVA molecules. The thermal properties of nanocomposite films have been investigated by thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC). The thermo-oxidative degradation temperature of PVA increases with the addition of In₂O₃ nanocrystals and the degree of crystallinity of the PVA matrix decreases in the presence of In₂O₃ nanocrystals in the nanocomposite films. The room temperature sensing characteristics of the naocomposite films have been studied for various gases, namely, H₂S, NH₃, CH₃, CO, and NO. The PVA–In₂O₃ nanocomposite films show maximum sensitivity for H₂S gas with fast response and reversibility. The response mechanism of the nanocomposite films to various gases is also proposed.

---

1. Introduction

The increasing concern to environmental protection and human health has generated great interest in efficient gas detection.¹ Hydrogen sulfide (H₂S) is a toxic, corrosive, and inflammable gas produced in sewage, coal mines, oil, and natural gas industries. It is used in large amounts in various chemical industries, research laboratories and as a process gas in the production of heavy water. H₂S can damage the human nerve and respiratory systems, causing people to lose consciousness or even die at very low concentrations (the occupational exposure limit for the gas is 10 ppm for 8 h exposure).² Therefore, a H₂S sensor that is sensitive and rapid in its response is needed for environmental and safety concerns.

Metal oxide semiconductor based gas sensors have been a subject of extensive research because of their potential application in detecting several inflammable, toxic, colorless and odorless gases.³ A large and fast change in the conductivity of oxide semiconductors because of annihilation and formation of oxygen vacancies in the presence of an oxidizing and reducing gas, respectively, forms the basis of the detection/conduction mechanism. The chemical sensing behavior is prominently a surface-dependent phenomenon and the surface-to-volume ratio of the sensing material significantly influences the sensing performance. Thus, metal oxide nanostructured materials are of great interest for gas-sensing applications due to their higher surface : volume ratio compared to bulk/thin film materials leading to better sensitivities.^4,5 However, it may not be always practical to use them as sensor materials as they suffer from high manufacturing expenses and the shaping and processing of these materials is often demanding.⁶

In₂O₃, is an n-type semiconductor in its nonstoichiometric form, due to intrinsic oxygen vacancies⁷ and has been extensively studied as a conductivity-based chemical sensor for the detection of various gases, viz., NO₂,⁸ HCHO,⁹ NH₃,¹⁰ and CO.¹¹ However, most of the sensors incorporating indium oxide require an elevated temperature (100–350 °C) for optimum operation which makes them highly power inefficient and reduces their long term stability. Also, reports on room temperature detection of H₂S gas by indium oxide nanostructures based sensors are sparse.^12–14 As a result, it is highly desirable to develop sensors that operate at room temperature with improved sensitivity and selectivity.

An alternative to nanostructured metal oxide materials for gas sensing applications are their composites with polymers. Polymers are flexible, lightweight materials and can be produced at a low cost. They can also be processed easily and can be shaped into thin films by various techniques such as dip-coating, spin coating, film casting and printing. Therefore, the drawbacks of using nano-sized metal oxides can be overcome by employing a polymer matrix to embed a relatively small content of metal oxide nanoparticles. The integration of inorganic nanoparticles into a polymer matrix allows both properties from inorganic nanoparticles and polymer to be combined/enhanced and novel functionalities can be generated in the polymer nanocomposites.^15,16 The properties of such nanocomposite materials depend not only on the properties of their constituents but also on their combined morphology and interfacial characteristics, which brings synergistic effects.

Polyvinyl alcohol (PVA) is a non-toxic, biodegradable, and water-soluble synthetic polymer. Due to its good film-forming nature, optical transparency and easy processability it has been extensively investigated as a host matrix for different kind of nanofillers.^17,18 To prepare nanocrystal–polymer nanocomposites, mainly two strategies are applied: (i) in situ, which comprise polymerization of the organic matrix around the inorganic nanoparticles¹⁹ and (ii) ex situ synthesis which includes direct incorporation of inorganic nanoparticles into the polymer matrix by mechanical blending, polymer melt intercalation or solvent processing.²⁰

In this paper we report the ex situ synthesis, characterization and gas-sensing properties of PVA based nanocomposite films consisting of In₂O₃ nanocrystals. The In₂O₃ nanoparticles used in this work were prepared by nonhydrolytic alcoholysis ester elimination reaction of indium acetate in the presence of oleic acid and oleyl alcohol at 220 °C. The nanocomposites of In₂O₃ nanoparticles in PVA were prepared by sonication-assisted dispersion followed by solvent casting. The gas-sensing properties of the PVA–In₂O₃ nanocomposite films have been explored for various gases at room temperature. The nanocomposite films show maximum sensitivity for H₂S gas with fast response and reversibility. To the best of our knowledge, ours is the first report on gas sensing of H₂S by PVA–In₂O₃ nanocomposite films at room temperature.

2. Experimental details

2.1 Materials

Indium acetate (99.99%, Aldrich), oleic acid (97%, Acros), oleyl alcohol (85%, Aldrich) and polyvinyl alcohol (PVA) (av. M_w = 14 000, Sigma) were used as received.

2.2 Synthesis of In₂O₃ nanocrystals

In₂O₃ nanocrystals were prepared by a nonhydrolytic alcoholysis ester elimination reaction of indium acetate. In a typical preparation, indium acetate (0.5 g, 1.71 mmol) was taken in a 100 ml three-necked flask with 40 ml (∼120 mmol) oleyl alcohol and 2.5 ml (∼15 mmol) of oleic acid. The reaction flask was evacuated to a vacuum level of 2 mbar for 30 min at 100 °C. The reaction system was then heated to 220 °C under argon flow and maintained at this temperature for 4 h. The solution was then cooled to 60 °C, and methanol was added to precipitate the nanocrystals. The off-white nanocrystals recovered by centrifugation, were dispersed in toluene and precipitated by addition of methanol. The redispersion/precipitation route was repeated twice to remove any unreacted precursors and excess oleic acid/oleyl alcohol.

2.3 Synthesis of polymer–In₂O₃ nanocomposite films

Pristine PVA and PVA–In₂O₃ nanocomposite films were prepared by a solvent-casting technique. To prepare composite films, a 5% (w/w) solution of PVA was prepared in nanopure water (conductivity = 0.6 μS cm⁻¹). The PVA was slowly added to water with continuous stirring and the mixture was later autoclaved at 110–115 °C for 10 min. To the clear solution so obtained, appropriate amount of In₂O₃ nanocrystals was added with continuous stirring followed by sonication assisted dispersion. The solution so obtained was poured into smooth polystyrene dishes and left in a dust free chamber at room temperature for four days to get nanocomposite films on evaporation of the solvent. The thickness of the produced films was ∼90 μm. PVA–In₂O₃ nanocomposite films consisting of 1 wt% In₂O₃ and 5 wt% In₂O₃ hence forth mentioned as PVA-IO1, and PVA-IO5, respectively, were prepared.

2.4 Characterization

Phase purity and structure of the In₂O₃ nanocrystals and the PVA–In₂O₃ nanocomposite films were determined by X-ray powder diffraction (XRD) data, which were collected on a Philips X'Pert pro X-ray diffractometer using Cu–Kα radiation (λ = 1.5418 Å) at 40 kV and 30 mA. The average crystallite size was calculated from the diffraction line width based on the Scherrer equation (d = 0.9λ/B cos θ, where λ is the wavelength (in Å) of X-rays, θ is the Bragg angle, and B is the half-maximum line width. Transmission electron micrographs (TEM) for the In₂O₃ nanocrystals were obtained using a FEI-Tecnai G-20 microscope operating at 200 keV. The TEM specimens were prepared by dispersing the samples in ethanol, and placing a drop of the dispersion on a Cu TEM grid covered with a holey carbon film, which was then dried. FTIR spectra of pristine PVA film and nanocomposite films were recorded in attenuation reflection mode (ATR) using a Bruker Alpha FTIR Spectrometer, in the range of 500–4000 cm⁻¹. Micro-Raman spectra of the nanocrystals and the nanocomposite films were obtained on a Horiba Jobin Yvon Labram-HR make spectrometer using an Ar ion laser (514.5 nm). The thermal characteristics of pristine PVA and nanocomposite film samples were determined by TGA and DSC. The thermogravimetric analysis was carried out using a Setaram Setsys thermogravimetry analyzer under oxygen, at a flow rate 16 ml min⁻¹. The heating rate in both cases was 10 °C min⁻¹. DSC measurements of pristine PVA films and PVA–In₂O₃ nanocomposite films were performed on a Mettler Toledo instrument (Model DSC 823^e) in the temperature range 30–250 °C. The heating rate as well as cooling rate was 20 °C min⁻¹.

2.5 Gas-sensing measurements

In order to measure the gas sensing characteristics of the films, the electrical contacts were prepared by first thermally evaporating two gold pads (120 nm thick with 1 mm spacing) and then attaching silver wires to them by silver paint. A change in resistance of the film as a function of time (response curve) was recorded at room temperature for 10 ppm concentration of different gases (i.e. H₂S, NH₃, CH₄, CO, and NO) which were commercially procured. The response curves were recorded using a static gas testing setup, which is described elsewhere.²¹ Briefly, the sensor was mounted upside down in a stainless steel test chamber (volume: 250 cm³). A desired concentration of the test gas in the chamber is achieved by injecting a known quantity of gas using a micro-syringe. The response data was acquired by using a personal computer equipped with Labview software. Once a steady state was achieved, recovery of sensors was recorded by exposing the sensors to air, which is achieved by opening the lid of the chamber. From the response curves, the sensitivity (S) was calculated using the relation:

S = R_a/R_g × 100%

where, R_a and R_g are resistances in air and test gas, respectively. Response and recovery times were defined as the time needed for 90% of total resistance change on exposure to gas and air, respectively.

3. Results and discussion

3.1 Synthesis

In₂O₃ nanocrystals have been conveniently synthesized by a nonhydrolytic alcoholysis route based on the well-known ester elimination reaction that involves the nucleophilic attack of the hydroxyl group of alcohol on the carbonyl carbon atom of metal acetate (Scheme 1).²² Pristine PVA and PVA–In₂O₃ nanocomposite films with two different fractions (1% and 5% w/w) of In₂O₃ nanocrystals, PVA-IO1 and PVA-IO5, respectively, were prepared by a solvent-cast method.

Scheme 1 Alcoholysis ester elimination reaction between indium carboxylate and alcohol.

3.2 Morphology and structure

X-ray diffraction (XRD) patterns (Fig. 1a), and transmission electron microscopy (TEM) images (Fig. 1b) clearly identify the In₂O₃ nanocrystals obtained in this work as highly crystalline with cubic bixbyite structure without the presence of any other impurity phase. In the TEM images, the particle sizes are typically in the range 10–16 nm, which is in agreement with the mean crystal size value of 13 nm estimated by Scherrer's equation of the (222) peak from the X-ray diffractogram. The well-resolved (321) lattice planes with a d spacing of 2.67 Å are also denoted in the high-resolution TEM (HRTEM) micrograph Fig. 1c. The selected area electron diffraction (SAED) pattern of the nanocrystals shown in Fig. 1d proves the high crystallinity of the nanocrystals.

Fig. 1 (a) XRD pattern of In₂O₃ nanocrystals (b) low-resolution TEM images of In₂O₃ nanocrystals (c) HRTEM image of a single In₂O₃ nanocrystal. The lattice spacing of 2.67 Å between adjacent lattice planes in the image corresponds to the distance between two (321) crystal planes. (d) SAED pattern for In₂O₃ nanocrystals.

PVA is known to be a partially crystalline polymer and the diffraction peaks at 2θ = 19.4°, and 40.4° correspond to the PVA crystalline phase.^23,24 The crystallinity of PVA results from the strong intermolecular interaction between the PVA chains due to intermolecular hydrogen bonding. Fig. 2 shows the XRD patterns of pristine PVA film and PVA containing 5% weight loading of In₂O₃ (PVA-IO5). In pristine PVA film, broad diffraction peaks at 2θ = 19.5, 22.5 and 40.5° are observed. The peaks at 19.5° and 22.5° correspond well to the typical reflection of the (1, 0, 1) and (1, 0, −1) planes of the semicrystalline PVA, respectively.²³ The peak at 40.5° is assigned to the (2, 2, 0) plane of the PVA.^25,26 In the X-ray diffractogram of composite film, besides the diffraction peaks of PVA, peaks at 2θ = 30.58, 35.46, 37.72, 39.81, 41.84, 45.64, 51.03, 55.94 and 60.69° correspond to the (2, 2, 2), (4, 0, 0), (4, 1, 1), (4, 2, 0), (3, 3, 2), (4, 3, 1), (4, 4, 0), (6, 1, 1), and (6, 2, 2) planes of the cubic bixbyite In₂O₃, respectively (JCPDS no. 88-2160). It is obvious that there is no significant effect on the general shape of the XRD pattern after incorporation of In₂O₃ nanocrystals. The pattern confirms that the In₂O₃ particles are embedded in the polymer matrix. Fig. S1† in the supplementary data shows the XRD pattern for PVA-IO1 sample.

Fig. 2 XRD patterns of pristine (a) PVA and (b) PVA-IO5 nanocomposite films. * Indicates peaks due to PVA.

3.3 FTIR studies

To investigate the interaction between PVA and In₂O₃ nanocrystals in the nanocomposite films, FTIR spectra have been measured. Fig. 3a shows the FTIR spectra of the pristine PVA and nanocomposite composite film, PVA-IO5, respectively. All the characteristic bands corresponding to PVA could be observed in a pristine PVA film.²⁷ In the spectrum of PVA, there is a broad and strong absorption at 3000–3600 cm⁻¹, peaking at 3273 cm⁻¹, which is due to the symmetrical stretching vibration of O–H from the intermolecular and intramolecular hydrogen bonds. The peaks at 2800–3000 cm⁻¹ are assigned to C–H stretching vibrations from the –CH₂–. A strong band at 1084 cm⁻¹ is assigned to stretching vibration of C–O in the C–O–H groups. The bands due to bending and wagging vibrations of CH₂ groups appear at 1414 and 1325 cm⁻¹, respectively, while C–H wagging is seen at 1263 cm⁻¹. The symmetric C–C stretching vibration which is an indication of the presence of crystalline region in PVA²⁸ is characterized by the band at 1142 cm⁻¹ and this band is observed in FTIR spectra of both pristine PVA and PVA-IO5 nanocomposite films. In case of PVA–In₂O₃ composite films, the frequency of the symmetrical stretching vibration of hydroxyl group moves to 3250 cm⁻¹, which is lower than that in a pristine PVA film, indicating slightly reduced hydrogen bonding among the PVA chains,²⁹ the intensity of the band is also reduced. When In₂O₃ nanocrystals are dispersed in the PVA matrix, they may form hydrogen bonds with PVA because of presence of hydroxyl groups on their surface, thus, reducing the hydrogen bonding among PVA molecules.

Fig. 3 FTIR spectra of pristine (a) PVA and PVA-IO5 nanocomposite films and (b) room temperature micro-Raman spectra for the PVA-IO5 nanocomposite film.

A new intense broad band centered at ∼590 cm⁻¹ in nanocomposite films is assigned to the In–O stretching vibration.³⁰ FTIR results confirm the incorporation of In₂O₃ nanocrystals in the PVA matrix and their interaction with PVA molecules, most likely due to interaction between surface hydroxyl groups on nanocrystals with –OH groups of PVA.

3.4 Raman spectroscopy

More evidence of the existence of In₂O₃ in the nanocomposite films is provided by Raman studies carried out on the PVA-IO5 sample. The room temperature micro-Raman spectrum of the nanocomposite film is shown in Fig. 3b. The Raman bands observed at 486, 520 and 631 cm⁻¹ are consistent with Raman active modes of the bcc In₂O₃ structure.³¹ The peaks at 486 and 631 cm⁻¹ are usually interpreted as stretching vibration υ (InO₆) of InO₆ octahedrons. The room temperature micro-Raman spectrum of In₂O₃ nanocrystals is shown in Fig. S2† of the supplementary information and expected vibrational modes at 301, 360, 490 and 624 cm⁻¹ corresponding to phonons associated with the bcc-structured indium oxide are observed.

3.5 Thermal characterization of PVA–In₂O₃ nanocomposites

The influence of incorporation of In₂O₃ nanocrystals on the thermal properties of PVA matrix was examined by TGA and DSC analysis. The thermo-oxidative stability of the PVA–In₂O₃ nanocomposite film was compared with the thermo-oxidative stability of a pristine PVA film. The TG and DTG curves for pristine PVA and PVA–In₂O₃ nanocomposite films recorded under a flowing oxygen atmosphere are shown in Fig. 4. As seen from Fig. 4a, the thermo-oxidation of pristine PVA occurs in six distinct steps. The % weight loss observed for each step along with corresponding temperature range is given in Table 1. The weight loss that emerges around 100 °C is a consequence of losing physically absorbed moisture. The mass loss observed in the temperature range 180–250 °C (∼6.8%) is due to partial dehydration of PVA chains which generates polyenes structure.^32,33 The next three thermal degradation steps which occur consecutively in the temperature range 254–449 °C (total mass loss ∼66%) are due to heating arrangement of the polyenes structure to polyaromatic form.³³ The decomposition of polyenes results in formation of macroradicals which further decompose to form cis and trans derivatives. The latter can form polyconjugated aromatic structures as a result of intramolecular cyclization and condensation reactions according to the Diels–Alder mechanism. The final step occurs between 450 and 545 °C and is due to thermo-oxidation of carbonized residue.³⁴

Fig. 4 TG, DTG and time temperature graph of (a) PVA and (b) PVA-IO5 nanocomposite films.

Table 1 TGA data for PVA and PVA-In₂O₃ nanocomposite films

Sample	Temperature range (^oC)	% Weight loss	Plausible thermal degradation step
PVA film	46–144	4.1	Removal of physically absorbed moisture
	180–250	6.8	Partial dehydration of PVA chains and formation of polyenes
	254–387	39.3	Decomposition and rearrangement of polyenes
	387–417	13.1	Chain-scission reactions, side-reactions and cyclization reactions (Diels–Alder intra- and intermolecular cyclization followed by dehydrogenation with aromatization)
	418–449	14.3	Cyclization and condensation processes of polyaromatic structures
	450–545	22.6	Oxidation of carbonized residue
PVA–In2O3 composite film (PVA-IO5)	48–148	5.96	Removal of physically absorbed moisture
	195-260	7.6	Partial dehydration of PVA chains and formation of polyenes
	264–379	34	Chain-scission reactions, side-reactions and cyclization reactions (Diels–Alder intra- and intermolecular cyclization followed by dehydrogenation with aromatization)
	380–411	11.61	Cyclization and condensation processes of polyaromatic structures
	412–503	33.1	Oxidation of carbonized residue

---

The TG data for the PVA–In₂O₃ (PVA-IO5) nanocomposite films is summarized in Fig. 4b and Table 1. It is clear from the data in figure and table that the presence of In₂O₃ nanocrystals in the PVA matrix did not significantly alter the degradation mechanism but did affect thermal stability. The increase in thermal stability of the composite film is seen in an increase of thermal degradation onset temperature during step 1 (temperature range 195–260 °C; ∼7.6% mass loss) and step 2 (temperature range 264–379 °C, ∼34% mass loss) by ∼15 °C and 10 °C respectively. The improved thermal stability may be assumed to be due to interactions between In₂O₃ nanocrystals and PVA. The interactions between In₂O₃ nanocrystals and PVA may suppress the chain-transfer reactions, slow the degradation process, and limit the motions of polymer chains, thereby increasing the decomposition temperature.³⁵ Further, a sudden mass loss, which is exothermic in nature, is seen in the temperature range 415–500 °C. This is also reflected in a sudden increase in temperature in the time–temperature curve. The step is due to thermo-oxidation of carbonized residue and is probably being catalyzed by In₂O₃ nanocrystals. The thermo-oxidation of the PVA-IO1 sample also shows similar behavior and TG data is shown in Fig. S3† of the supplementary information.

DSC curves of pristine PVA and PVA–In₂O₃ nanocomposite films (PVA-IO1 and PVA-IO5) during the heating cycle in the temperature range 25–250 °C are shown in Fig. 5a. Inset in the figure shows the DSC thermograms recorded during cooling cycle. Both pristine PVA as well as PVA–In₂O₃ nanocomposite films exhibit similar thermal response during heating cycle with three distinct characteristic features observed as a function of temperature. The first endothermic step observed around 315 K (∼42 °C) for pristine PVA and PVA–IO1 nanocomposite films is attributed to the glass transition temperature (T_g). The glass transition temperature of PVA is known to occur at ∼86 °C (359 K) but occurs at about 48 °C (321 K) in the presence of moisture.³⁶ The presence of In₂O₃ nanocrystals in the PVA matrix slightly increases the T_g value of polymer by ∼5 °C for the PVA-IO5 sample, along with a higher energy requirement as seen from the clear peak like shape of glass transition, however, no significant shift in T_g of PVA is observed for the PVA-IO1 sample. Glass transition process is considered to be affected by molecular packing, chain rigidity and linearity. The slight increase in T_g with incorporation of In₂O₃ nanocrystals may be because of confinement effects and weak interactions between the polymer and the filler In₂O₃ nanocrystals. These intercalations lead to restricted segmental mobility of the polymer chains. The observed increase in T_g of the PVA matrix is very similar to the results observed in literature.^37,38

Fig. 5 (a) DSC thermograms of pristine PVA and PVA–In₂O₃ nanocomposite films recorded during a heating cycle. Inset shows the cooling cycle behavior for the same materials. (b) DSC thermograms of PVA-IO1 nanocomposite films recorded during successive heating cycles. Inset shows the cooling cycle behavior for the same material.

A broad endothermic peak, in the temperature range 325–420 K with its maxima around 373 K observed for pristine PVA as well as PVA-IO1 is attributed to removal of free (moisture) as well as hydrogen bonded water and corresponding relaxation of intermolecular hydrogen bonds between the hydroxyl groups of PVA.³⁹ The absence of any such peak (exotherm) in the thermograms recorded during the cooling cycle confirms this relaxation phenomenon associated with the removal of water molecules. The third endothermic peak in the DSC thermograms with peak temperatures around 496 K, 493 K and 461 K for pristine PVA, PVA-IO1 and PVA-IO5 samples, respectively, corresponds to the melting temperature (T_m) of PVA. It is clear that PVA–In₂O₃ nanocomposite films melt at lower temperature as compared to pristine PVA and the effect is more pronounced with an increasing concentration of In₂O₃ nanocrystals. Also, the melting occurs over a broad temperature range for nanocomposite films compared to pristine PVA. A similar trend of a decrease in the melting temperature has been observed in PVA–CuO composites,^40a Ce(III) doped PVA,^40b Mg–Al layered double hydroxide–PVA composites,^40c and ZnO–PVA/PEO composites.³⁶ The decrease in T_m might be related to a decrease in the crystallinity of the sample. It is well known that PVA crystallizes as a result of folding of polymer chains, which themselves are arranged in a parallel fashion to each other. The incorporation of indium oxide nanocrystals in a semi-crystalline PVA matrix may cause adsorption of polymer chains over the nanocrystals surface, which in turn decreases the mobility of these polymer chains as well as crystal perfection. Consequently, heat resistance of the nanocomposite film also decreases, leading to lower melting temperatures.³⁹ The crystallinity of PVA in the nanocomposite films may also decrease because In₂O₃ nanocrystals prevent the interactions between the main chain and the formation of hydrogen bonds (see FTIR discussion and Fig. 3b).³⁶

In order to understand the effect of In₂O₃ nanocrystals on the structural characteristics of the PVA network, the degree of crystallinity of PVA was calculated from the enthalpies of fusion using the following equation:

X_c = [ΔH_f/(w·ΔH_f,100)] × 100 (1)

where ΔH_f and ΔH_f,100 are the heat of fusion as measured for the sample and the average heat of fusion of 100% crystalline PVA (∼138.6 J g⁻¹),⁴¹ respectively and w is the weight fraction of PVA in the composite. Fig. 5b shows the DSC thermograms recorded during successive heating–cooling cycles for the PVA-IO1 sample. The inset shows the DSC curves recorded during the cooling cycles. Table 2 lists the experimentally found values of X_c along with other parameters such as glass transition and melting temperature for the pristine and nanocomposite films (PVA-IO1 and PVA-IO5) and the effect of thermal cycling on the parameters deduced. It is evident from the decreasing trend of degree of crystallinity that incorporation of In₂O₃ nanocrystals into the PVA matrix leads to increasing amorphization. The broad temperature range over which the melting occurs in case of nanocomposite films also confirms this.

Table 2 Effect of successive heating–cooling cycles on the thermal properties of PVA–In₂O₃ nanocomposite films

Sample	Heating cycle	Tg (peak)	Tm (peak)	X c (%)
Pristine PVA		315 K	495.5 K	48.5
PVA-IO1	First heating	314.5 K	493.2 K	48.1
	Second heating	348 K	489.6 K	37.8
	Third heating	348 K	487.3 K	32.6
PVA-IO5	First heating	320.6 K	461 K	27.5
	Second heating	343.1 K	439.5 K	15.6

---

A shift of glass transition temperature (T_g) towards higher values for the 2nd and successive heating of samples is evidence of a moisture induced lowering of the ‘T_g’ of PVA based composites.³⁶ In the case of nanocomposite films, the melting occurs at lower temperatures upon successive heating cycles with an associated decrease in the degree of crystallization. A similar trend is observed during the cooling cycle DSC runs. This could be explained due to increased amorphization of the system upon successive heating–cooling cycles resulting into reduced thermal stability towards melting.

4. Gas-sensing

Response time, recovery time and selectivity are the important parameters for a gas sensor. Thus, a change in resistance of the PVA and PVA–In₂O₃ nanocomposite films (PVA-IO5) as a function of time (response curve) was recorded at room temperature for 10 ppm concentration of different gases, (i.e. H₂S, NH₃, CH₄, CO, and NO etc.) which were commercially procured. The PVA film was found to be insensitive to all the gases studied (Fig. 6a). On the other hand, the PVA-IO5 sample shows maximum sensitivity for H₂S gas and is almost insensitive to all the other gases studied (S ∼1) (Fig. 6b). Further to these results, the room temperature sensitivity of the PVA–In₂O₃ nanocomposite film towards different concentration of hydrogen sulfide gas was studied by measuring the change in resistance with time. Fig. 6c plots the sensitivity values observed for six different concentrations of H₂S ranging from 500 ppb to 10 ppm as a function of time. It is clear from the figure that the nanocomposite films, PVA-IO5 could sense the H₂S gas at the 500 ppb level. Also, the sensitivity increases as the concentration of gas increases. In all, five different sensors based on PVA-IO5 films were tested and it was found that the sensors produce repeatable responses of the same magnitude with good baseline stability, confirming the reproducibility. A typical response curve of a PVA-IO5 nanocomposite film for 10 ppm H₂S gas is shown in Fig. 6d. The film exhibited S = 3 with a response time of ∼25 s. It can also be concluded from the figure that full recovery of base resistivity is achieved in ∼410 s and the values of S is reproducible in repeated exposure cycle. Fig. 6e shows that S varies linearly with H₂S concentration saturating at 10 ppm. The composite films were stable in atmospheric conditions (temperature 30–38 °C and relative humidity 70–85%) for several months. These features indicate that nanocomposite films can be used as room temperature operating H₂S sensors in the range of 1–10 ppm.

Fig. 6 (a) Response curve for pristine PVA film on exposure to different gases, (b) bar chart showing the S value of nanocomposite film for 10 ppm of various gases, (c) response curve for the composite film on exposure to different concentrations of H₂S gas, (d) response curve of nanocomposite film on exposure to 10 ppm of H₂S gas. (e) Variation in S of composite film as a function of H₂S concentration.

4.1 Gas sensing mechanism

The mechanism of electric resistivity change in polymer materials with encapsulated semiconducting metal oxide nanoparticles in the presence of gas molecules is not very well understood. It is known that the resistivity of these materials is affected by the concentration and size of the nanoparticles (fillers).⁴² In the case of insulating polymers with conducting fillers, the resistivity of a material will change only when the fillers form a continuous path through the material, from one side to the other. The term ‘percolation threshold' is used for the lowest filler concentration at which this happens. A theoretically predicted value of percolation threshold for randomly distributed nonoverlapping hard spheres, which touch each other at the percolation threshold is 16 vol%.⁴³ However, polymer nanocomposites with very low critical filler fractions have been reported in the literature,^44,45 even down to a critical volume fraction of 0.03 vol%.^44a Based on the investigations by various groups, it is generally accepted that the properties of the fillers, such as size, aspect ratio, and surface groups on the filler particles, the aggregate/agglomerate structure of the filler particles in the matrix before/during final processing, the properties of the matrix including viscosity and crystallinity, and the interfacial energy between matrix and filler determine the percolation threshold of a polymer conducting composite.⁴⁶

Godovsky and coworkers⁴⁷ have studied the sensor behavior of PVA–CuS (particle size 12 nm) nanocomposites towards water vapor. They stated that the dependence of resistivity of CuS–poly-vinylacetates (12 nm size of filler particles) on water pressure could be due to the variation in dielectric constant of the polymer upon adsorption of polar molecules, resulting in the formation of a chain-like conducting cluster. The authors argue that there is the existence of hopping conductivity between the nanoparticles through the polymer in a range of concentrations close to the percolation threshold, which shows exponential dependence on dielectric permeability of the medium. This means that even small changes in polymer matrix dielectric permeability caused by adsorption of polar gas molecules will lead to an exponential change in the composite conductivity.

The change in resistivity in the PVA–In₂O₃ nanocomposite films studied in this work in the presence of gases could be due to the above mechanism. As the PVA film did not sense any gas the change in resistivity cannot be due to the matrix. In₂O₃ is an n-type semiconductor and the resistance of an n-type semiconductor is known to decrease on exposure to a reductive gas and increase on exposure to an oxidizing gas.¹² In our case, it was found that the resistance of the nanocomposite films always shows a decrease whether it is exposed to reducing gas (H₂S) or oxidizing gas (NO) (graph not shown). Thus, the change in resistivity is not due to chemisorption of gases on the surface of In₂O₃ nanocrystals, otherwise we should have observed an increase in resistance on exposing the composite film to an oxidizing gas like NO. The percolation threshold in this case could be achieved at a much lower concentration of nanocrystals in the polymer matrix as the percolation threshold tends to decrease with a reduction in particle size.

Conclusions

PVA based nanocomposite films with In₂O₃ (1 and 5 wt% loading) nanocrystals as filler have been successfully prepared. The incorporation of In₂O₃ nanocrystals into the PVA matrix causes induced changes in thermal properties and crystallinity of the PVA. The changes were discussed in terms of interaction between nanofillers and PVA molecules and this is supported by FTIR data. The gas sensor fabricated using PVA–In₂O₃ nanocomposite films show repeatable and good response towards H₂S gas with short response and recovery times at room temperature. Due to room temperature operation, the nanocomposite gas sensors may prove to be useful for environmental and industrial applications after further development.

Acknowledgements

The authors thank Dr R. Mishra of the Chemistry Division, BARC for providing TG data.

References

1. Y. Wang, F. H. Kong, B. L. Zhu, S. R. Wang, S. H. Wu and W. P. Huang, Mater. Sci. Eng., B, 2007, 140, 98.
2. J. Chou, Hazardous gas monitors, Mcgraw Hill,New York, 2000.
3. (a) M. Kugishima, K. Shimanoe and N. Yamazoe, Sens. Actuators, B, 2006, 118, 171; (b) V. N. Singh, B. R. Mehta, R. K. Joshi, F. E. Kruis and S. M. Shivaprasad, Sens. Actuators, B, 2007, 125, 482.
4. E. Comini, Anal. Chim. Acta, 2006, 568, 28.
5. D. H. Zhang, Z. Q. Liu, C. Li, T. Tang, X. L. Liu, S. Han, B. Lei and C. Zhou, Nano Lett., 2004, 4, 1919.
6. H. Althues, J. Henle and S. Kaskel, Chem. Soc. Rev., 2007, 36, 1454.
7. J. R. Bellingham, A. P. Mackenzie and W. A. Phillips, Appl. Phys. Lett., 1991, 58, 2506.
8. P. Xu, Z. Cheng, Q. Pan, J. Xu, Q. Xiang, W. Yu and Y. Chu, Sens. Actuators, B, 2008, 130, 802.
9. T. Chen, Q. J. Liu, Z. L. Zhou and Y. D. Wang, Sens. Actuators, B, 2008, 131, 301.
10. N. Du, H. Zhang, B. D. Chen, X. Y. Ma, Z. H. Liu, J. B. Wu and D. R. Yang, Adv. Mater., 2007, 19, 1641.
11. K. I. Choi, H. R. Kim and J. H. Lee, Sens. Actuators, B, 2009, 138, 497.
12. M. Kaur, N. Jain, K. Sharma, S. Bhattacharya, M. Roy, A. K. Tyagi, S. K. Gupta and J. V. Yakhmi, Sens. Actuators, B, 2008, 133, 456.
13. Z. Zeng, K. Wang, Z. Zhang, J. Chen and W. Zhou, Nanotechnology, 2009, 20, 045503.
14. K. Yao, D. Caruntu, Z. Zeng, J. Chen, C. J. O'Connor and W. Zhou, J. Phys. Chem. C, 2009, 113, 14812.
15. C. Sanchez, B. Julián, P. Belleville and M. Popall, J. Mater. Chem., 2005, 15, 3559.
16. D. Koziej, F. Fischer, N. Kränzlin, W. R. Caseri and M. Niederberger, ACS Appl. Mater. Interfaces, 2009, 1, 1097.
17. R. V. Kumar, Y. Koltypin, Y. S. Cohen, D. Aurbach, O. Palchik, I. Felner and A. Gedanken, J. Mater. Chem., 2000, 10, 1125.
18. (a) D. Y. Godovsky, Adv. Polym. Sci., 2000, 153, 165; (b) X. F. Qian, J. Jin, J. C. Huang, Y. F. Yang, X. X. Guo and Z. K. Zhu, Mater. Chem. Phys., 2001, 68, 95; (c) Z. H. Mbhele, M. G. Salemane, C. G. C. E. van Sittert, J. M. Nedeljković, V. Djoković and A. S. Luyt, Chem. Mater., 2003, 15, 5019.
19. (a) S. M. Khaled, R. Sui, P. A. Charpentier and A. S. Rizkalla, Langmuir, 2007, 23, 3988; (b) S. Scholz and S. Kaskel, J. Colloid Interface Sci., 2008, 323, 84.
20. (a) N. Nakayama and T. Hayashi, J. Appl. Polym. Sci., 2007, 105, 3662; (b) J. L. H. Chau, Y. M. Lin, A. K. Li, W. F. Su, K. S. Chang, S. L. C. Hsu and T. L. Li, Mater. Lett., 2007, 61, 2908.
21. (a) M. Kaur, D. K. Aswal and J. V. Yakhmi in Science and technology of chemiresistive gas sensors, ed. D. K. Aswal and S. K. Gupta, Nova Science Publisher, NY, USA, 2007; (b) P. Kanitkar, M. Kaur, S. Sen, A. Joshi, V. Kumar, S. K. Gupta and J. V. Yakhmi, Int. J. Nanotechnol., 2010, 7, 883.
22. X. Zhong, Y. Feng, Y. Zhang, I. Lieberwirth and W. Knoll, Small, 2007, 3, 1194.
23. M. Watase and K. Nishinari, J. Polym. Sci., Part B: Polym. Phys. Ed., 1985, 23, 1803.
24. C. W. Bunn, Nature, 1948, 161, 929.
25. G. M. Kim, A. S. Asran, G. H. Michler, P. Simon and J. S. Kim, Bioinspir. Biomimetics, 2008, 3, 046003.
26. R. Ricciardi, F. Auriemma, C. De Rosa and F. Laupretre, Macromolecules, 2004, 37, 1921.
27. (a) H. Tadokoro, Bull. Chem. Soc. Jpn., 1959, 32, 1252; (b) S. Akhter, K. Allan, D. Buchanan, J. A. Cook, A. Campion and J. M. White, Appl. Surf. Sci., 1988, 35, 241.
28. H. Tadokoro, K. Kôzoi, S. Seki and I. Nitta, J. Polym. Sci., 1957, 26, 379.
29. (a) H. J. Salavagione, G. Martinez and M. A. Gomez, J. Mater. Chem., 2009, 19, 5027; (b) L. Jiang, X. P. Shen, J. L. Wu and K. C. Shen, J. Appl. Polym. Sci., 2010, 118, 275.
30. Y. Xie, Z. Shi and J. Liu, Adv. Mater. Res., 2011, 239–242, 322.
31. W. B. White and V. G. Keramidas, Spectrochim Acta, 1972, 28A, 501.
32. P. Budrugeac, J. Therm. Anal. Calorim., 2008, 92, 291.
33. J. W. Gilman, D. L. Van der Hart and T. Kashiwagi, Fire and polymers II: materials and test for hazard prevention, American Chemical Society, ACS Symposium Series 599, August 21–26, Washington, DC, 1994.
34. A. Y. Shaulov, S. M. Lomakin, T. S. Zarkhina, A. D. Rakhimkulov, N. G. Shilkina, Y. B. Muravlev and A. A. Berlin, Dokl. Phys. Chem., 2005, 403, 154.
35. A. Leszczynska, J. Njuguna, K. Pielichowski and J. R. Banerjee, Thermochim. Acta, 2007, 453, 75.
36. J. Lee, D. Bhattacharyya, A. J. Easteal and J. B. Metson, Curr. Appl. Phys., 2008, 8, 42.
37. Z. Peng, L. X. Kong and S. D. Li, J. Appl. Polym. Sci., 2005, 96, 1436.
38. W. Pan, H. Li and Y. Chen, Appl. Mech. Mater., 2010, 26–28, 127.
39. A. Radosavljević, D. Božanić, N. Bibić, M. Mitrić, Z. K. Popović and J. Nedeljković, J. Appl. Polym. Sci., 2012, 125, 1244.
40. (a) R. Vijaya Kumar, R. Elgamiel, Y. Diamant, A. Gedanken and J. Norwig, Langmuir, 2001, 17, 1406; (b) M. Abdelaziz, Phys. B, 2011, 406, 1300; (c) B. Ramaraj, S. K. Nayak and K. R. Yoon, J. Appl. Poly. Sci., 2010, 116, 1671.
41. Z. Peng, L. X. Kong and S. D. Li, Polymer, 2005, 46, 1949.
42. D. Y. Godovskii, Adv. Polym. Sci., 1995, 119, 281.
43. D. Stauffer and A. Aharony, Introduction to percolation theory, Taylor & Francis, London , 1992.
44. (a) D. van der Putten, J. T. Moonen, H. B. Brom, J. C. M. Brokken-Zijp and M. A. J. Michels, Phys. Rev. Lett., 1992, 69, 494; (b) L. J. Adriaanse, J. A. Reedijk, P. A. A. Teunissen, H. B. Brom, M. A. J. Michels and J. C. M. Brokken-Zijp, Phys. Rev. Lett., 1997, 78, 1755.
45. (a) L. Flandin, M. Verdier, B. Boutherin, Y. Brechet and J. Y. Cavaille, J. Polym. Sci., Part B: Polym. Phys., 1999, 37, 805; (b) M. P. Clingerman, E. H. Weber, J. A. King and K. H. Schulz, J. Appl. Polym. Sci., 2003, 88, 2280; (c) L. Jong, J. Polym. Sci., Part B: Polym. Phys., 2005, 43, 3503; (d) Z. Chen, J. C. M. Brokken-Zijp and M. A. J. Michels, J. Polym. Sci., Part B: Polym. Phys., 2006, 44, 33.
46. (a) L. Karasek and M. Sumita, J. Mater. Sci., 1996, 31, 281; (b) P. Banerjee and B. M. Mandal, Macromolecules, 1995, 28, 3940; (c) T. Slupkowski, Phys. Status Solidi A, 1984, 83, 329; (d) M. A. J. Michels, J. C. M. Brokken-Zijp, D. Van der Putten, J. T. Moonen and H. B. Brom, Phys. Rev. Lett., 1993, 70, 4161.
47. D. Y. Godovski, E. A. Koltypin, A. V. Volkov and M. A. Moskvina, Analyst, 1993, 118, 997.

---

Footnote

† Supplementary information provided: X-ray diffraction data for PVA-IO1 sample, micro-Raman data for In₂O₃ nanocrystals and TG data for PVA-IO1 sample.

---