High mechanical and pressure sensitive dielectric properties of graphene oxide doped PVA nanocomposites

Abstract

We report the significant enhancement in mechanical stiffness, Young’s modulus and tensile strength properties of graphene oxide (GO) doped poly(vinyl alcohol) (PVA) nanocomposites. The prepared nanocomposites exhibit a relatively high dielectric constant with a small loss factor and significant response to applied pressure. High mechanical strength is observed for the lower concentration of GO. Recognition to the applied pressure technique has been demonstrated to study the behavior of dielectric and electrical responses. In this study, a unique and sensitive change in dielectric and electrical properties was observed, which can be used to open up new design space for pressure sensing applications.

---

1. Introduction

In recent years, polymer-based nanocomposites have been intensively studied with the aim to obtain high-performance, lightweight materials. In order to achieve this purpose, carbon based materials (graphite nanoplatelets, graphene sheets, carbon nanotubes (CNT) and graphene oxide (GO)) have drawn intense excitement as conventional fillers, because of their enhancing mechanical and thermal properties.^1–4 Several polymeric composites have been employed as hosts for carbon nanofiber (CNF) and nanotube (CNT) filler, including polyolefins, poly(acrylates), epoxy, elastomers etc. Nanocomposites with GO have also been produced using hydrophobic polymers, including cellulose,² chitosan,³ polyamide⁴ and poly(methyl methacrylate) (PMMA).⁵ CNTs have been regarded as one of the most effective filler materials, but due to expensive multi-step methods used to prepare and purify, CNTs have limited production on an industrial scale and also poor dispersibility in water or organic solvents, which limits practical application. Graphene-based fillers are thus expected as a promising replacement or supplement to CNTs. Their mechanical strength is comparable to that of CNTs and their shapes may make them superior to nanotubes for some applications. Hence, it will be extremely interesting to fabricate composites with graphene materials to improve their mechanical properties.^1–6

GO is a functionalized graphene material, bears oxygen-containing functional groups on the basal planes and edges of graphene⁷ and some of its interesting properties are: (i) its good dispersibility in water, due to the presence of many oxygen functional groups, such as hydroxyls, epoxides, carboxyls and carbonyls, (ii) a substantial band gap for the presence of a large amount of sp³ hybridized carbon, making it a unique material from graphene and reduced graphene oxide and (iii) its 2D nanostructure and chemical functionality make it highly attractive in nanocomposites.⁸ On the other hand, poly(vinyl alcohol) (PVA) is a water-soluble synthetic polymer with high hydrophilicity, biocompatibility and non-toxicity. It has excellent film-forming, emulsifying and adhesive properties. The high capacity of PVA to form both intra- and inter-chain hydrogen bonds simultaneously makes it an interesting polymer which could interact with GO.

Stankovich et al. have shown that a graphene oxide doped polystyrene composite formed by complete exfoliation of graphite and a molecular-level dispersion route exhibits the highest electrical conductivity (1 S m⁻¹) for 2.5 wt% doping at room-temperature.⁹ In another study, Ramanathan et al., have reported the significant improvement in thermal properties (30 °C) for 1 wt% functionalized graphene sheets in poly(acrylonitrile) and poly(methylmethacrylate) polymers.¹⁰ The pressure dependent ac conductivity and complex impedance measurements of conducting polypyrrole was studied and pressure-sensitive frequency regions attributed to inter-chain hopping relaxation mechanism was reported.¹¹ In another study it has been reported that the degree of oxygen content of graphene oxide can be varied by varying KMnO₄ concentration for various applications.¹² S. M. Imran et al. have studied the carbon nanotube-based thermoplastic polyurethane–poly(methyl methacrylate) nanocomposites for pressure sensing applications.¹³

GO doped with polymer composites with significantly improved mechanical properties can be found in reported papers.¹⁴ The polymer matrix and GO are expected to bond to each other by weak intermolecular forces, and chemical bonding is rarely involved. If the reinforcing material in the composite could be dispersed on a molecular scale and interact with the matrix by chemical bonding, then significant improvements in the mechanical properties of the material or unexpected new properties could be attained. In view of this, various processing methods have been developed to fabricate graphene oxide/polymer nanocomposites, including in situ intercalative polymerization, solution mixing, melt blending and self-assembly. From an economic point of view, the solution mixing approach should be the most efficient and scalable method for producing graphene-filled nanocomposites.

In order to increase the mechanical properties of PVA, especially to form stable intermolecular hydrogen bonds between the hydroxyl groups on PVA chains, we prepared PVA-based nanocomposites filled with GO. Because GO contains abundant oxygen groups and PVA contains hydroxyl groups, a homogeneous dispersion of GO into PVA and strong interfacial adhesions between them can be achieved. A significant reinforcement by GO has been observed for these PVA/GO nanocomposites from the increases in tensile strength, Young’s modulus and elongation at break. Furthermore, we focused on the dielectric behavior of prepared composites at different temperatures and pressures. In addition, the effect of GO doping on the microstructural, optical, thermal properties of the prepared composites is also systematically investigated.

2. Experimental

2.1 Materials

PVA was purchased from SDFCL (sd fine-chem limited) Mumbai, India. The molecular weight of PVA is 1, 25 000 and the degree of saponification is 86–89%. Graphite, sodium nitrate and potassium permanganate were all supplied by Loba Chem, Mumbai, India and were used as received.

2.2 Preparation of GO

A modified Hummers method was utilized to synthesize the oxidized graphite powders.⁵ Typically, 1.25 g of natural graphite, 1.25 g of NaNO₃ and 3.25 g of KmnO₄ were slowly added to 25 mL of concentrated H₂SO₄ under vigorous stirring below 5 °C. The mixture was then stirred continuously for 1 h at 35 °C to oxidize the graphite. Afterward, 50 mL of water was added into the mixture and the temperature was increased to higher than 80 °C, and the suspension was maintained at 90 °C for 15 min. The mixture was then poured into 150 mL of deionised water, after that 10 mL of H₂O₂ was added into the suspension. After cooling to room temperature, the solid products were filtered, subsequently washed with 5% HCl aqueous solution and water, and dried to obtain graphene oxide.

2.3 Preparation of GO–PVA solid polymer composites

In a typical process, PVA polymer composites are prepared using the solution casting method. Initially, a known quantity of PVA was added to water and kept for 72 hours at room temperature for complete dissolution. Then the produced GO was dispersed in prepared PVA solution, at different (1, 2, 3, 4, 5 wt%) weight percentages. This solution was kept aside till it became viscous and then the obtained viscous solution was casted onto a clean glass plate and dried under vacuum at room temperature for 48 h to form thin films of 100–150 micron in thickness. The weight fractions of PVA–GO films in the composite can be determined by using the relation,where m_d and m_p is the weight of the dopant and polymer, respectively.

2.4 Characterization of prepared composites

Fourier transform infrared (FT-IR) spectra were measured using a Shimadzu-21 spectrometer in the absorption mode with a resolution of 4 cm⁻¹. The UV-Vis optical absorption spectra of the composite films are recorded using Shimadzu-1800 spectrophotometer in the wavelength range of 190–800 nm. Thermal gravimetric analysis (TGA) of the samples was carried out by using TA Instruments SDT Q600 with a heating rate of 10 °C min⁻¹ under nitrogen atmosphere. XRD measurements are performed using a Bruker D8 Advance X-ray diffractometer with Ni-filtered, Cu-K_α radiation of wavelength 1.5406 Å using a graphite monochromator. The scan is taken in the 2θ diffraction angle in the range from 5–50° with a scanning speed and step size of 1 deg min⁻¹ and 0.01° respectively. Mechanical properties of the 50 mm × 25 mm composite films are measured with the Lloyd LRX plus universal testing machine (UTM) equipped with a 5 kN load cell. The film samples in the form of strips of the size mentioned above are fixed in grips at a gauge length of 50 mm and the mechanical measurements are performed at room temperature with a constant rate of 25 mm min⁻¹. The stress in the form of load is applied to stretch the sample until it breaks. The graph of stress verses percentage strain gives the modulus, tensile strength, stiffness and elongation at fracture behaviors of the material. It should be noted that more than five samples of each material were tested from which the mean and standard deviation were calculated.

The impedance measurements were performed using a HIOKI-IM3570 high precession impedance analyzer interfaced to a computer at an applied voltage of 0.3 V and a frequency in the range of 4 Hz to 5 MHz and a temperature range of 303–343 K. Samples were mounted on the conductivity holder with stainless steel electrodes with a diameter of 1.76 cm under spring pressure. The value of bulk resistance (R_b) obtained from the measurement was used to calculate the conductivity (σ) using the equation σ = t/R_bA, where A is area of the electrode, t is the thickness of the film and bulk resistance can be retrieved from the intercept of the high frequency semi-circle or the low frequency spike on the Z-axis.

3. Results and discussion

Fig. 1 shows the FT-IR spectra of PVA and GO doped PVA composite films. For pure PVA, a strong broad band at 3447 cm⁻¹ is observed and is assigned to the O–H stretching vibration of hydroxyl groups of PVA. Similarly, the IR bands at 2926 cm⁻¹ (C–H asymmetric stretching), 2858 cm⁻¹ (C–H symmetric stretching), 1744 cm⁻¹ (C=O stretching) and 1632 cm⁻¹ (acetyl C=C), 1431 cm⁻¹ (CH₂) were detected and a sharp band at 1089 cm⁻¹ was assigned to the C–O stretching group present on the PVA backbone.

Fig. 1 FTIR spectra of pure and different concentrations of the GO doped PVA composite.

For the GO doped PVA composite film, a decrease in intensity around 3447 cm⁻¹ is observed and this peak is broadened and shifted to a lower wavenumber with the addition of GO confirming the presence of the free and associated hydroxyl groups in the polymer matrix. The shift of the peak from 3447 cm⁻¹ (pure PVA) to 3396 cm⁻¹ for 5 wt% GO doped composites is attributed to the increase in strength of hydrogen bond interactions existing between the oxygen-containing functional groups (carboxyls, five-membered-ring lactols, and hydroxyls (on both the basal plane and the edge hydroxyls)) of GO and the hydroxyl groups on the PVA molecular chains and also, with the increase in GO doping concentration, there is an increase in the formation of CTC with OH groups of the PVA matrix.

The band corresponding to the C–H asymmetric stretching vibration is observed at 2928 cm⁻¹ and the C–H symmetric stretching vibration at 2853 cm⁻¹.¹⁰ The intensity of the stretching peak around 1733 cm⁻¹ with higher intensity in the PVA/GO spectrum corresponds to the carboxyl groups adjacent to benzene groups, indicating on one hand that PVA has indeed been covalently attached to GO and after introduction of GO, the carbonyl peak shifted from 1733 to 1722 cm⁻¹, indicating that a strong hydrogen bond exists between GO and PVA, which not only promotes the level of GO dispersion but also results in improved properties of the composite.^14,15

The presence of the band at 1632 cm⁻¹ is evident by the presence of a small amount of adsorbed water in the PVA–GO composite and the peak at 1652 cm⁻¹ (C=C stretching vibration of PVA) shifts to 1654 cm⁻¹ for 5 wt% GO; this is attributed to the interaction with the unoxidized graphitic domains and the presence of vibrational bands at 1433 cm⁻¹ is assigned to the deformation vibrations of C–OH groups. There is no peak position shift at 1433 cm⁻¹ after GO loading, indicating that there is no covalent bonding between PVA chains and GO, but only simple π–π stacking and a weak absorption peak at 1356 cm⁻¹ was assigned to the stretching vibrations of carboxyls. Symmetric stretching of epoxy (C–O–C at 1261 cm⁻¹), and alkoxy (C–O at 1029 cm⁻¹) groups indicate that the surface of GO is decorated by oxygenic groups and an aromatic C–H peak at 832 cm⁻¹. A band at about 800 cm⁻¹ could be attributed to the C–H out-of-plane bending vibrations, while the many low-intensity peaks ranging from 780 to 580 cm⁻¹ can be assigned to the vibrations of the C–H bonds in the benzene rings.¹⁶

Optical measurements were performed on the pure PVA and GO doped PVA films of different concentrations are shown in Fig. 2. The absorption band at 196 nm in PVA is assigned to n–π* transitions and around 208 nm indicates the presence of unsaturated bonds (C=O and/or C=C) mainly in the tail–head of the polymer PVA. The peak at 193 nm has shifted to 212 nm for 5 wt% GO doped composites. The above red shifts may be attributed to the decrease of planarity of GO due to hydrogen bonding with the high-molecular-weight of the PVA chain, resulting in less delocalization of the π–π* transition in the conjugated system and a shoulder at about 247 nm, which could be assigned to n–π* transitions of C=O bonds, respectively and also gradual restoration of the π-conjugation network within GO. As the GO concentration increases, the inter/intra hydrogen bonding increases and hence absorption.^8,17

Fig. 2 (i) UV absorption spectra of the GO doped PVA films; (ii) variation of (αhν)^1/2 with photon energy (hν) of GO doped PVA composites.

The band gap was calculated using spectrophotometry where the absorption coefficient is calculated from the following relation:

where T is transmittance measured directly from the instrument, t is film thickness and (1 − R) ∼ 1 is the normal incidence.

In order to determine the optical band gap, a graph of (αhν)^1/2 is plotted as a function of hν (photon energy), where α is the absorption coefficient as shown in Fig. 2(ii). The absorption coefficient decreases slightly with the addition of GO and this fact may be attributed to the increase in the samples reflectivity. Extrapolating the linear region of the curve to the x-axis gives the optical band gap and the optical energy band gap (E_g) for the pure as well as doped PVA films decreases with the increase in GO concentration due to the presence of the large surface area of graphene oxide and also maximum change is observed at low loading where maximum interaction takes place between PVA and GO.^18,19

The DSC thermograms of GO doped PVA composites from 50 °C to 350 °C are shown in Fig. 3. Since PVA is a semicrystalline polymer, its physical and microstructural properties are strongly dependent on the degree of crystallinity. The glass transition region was analyzed in order to investigate the influence of mobility of the chains with temperature. Significant changes are observed in the glass transition temperature (T_g), the melting temperature (T_m) and the degradation temperature (T_d) of PVA depending on the filler content.¹⁵

Fig. 3 DSC curves of GO doped PVA nanocomposites.

From Fig. 3, it is observed that the T_g increases from 68 °C of pure PVA to 76 °C for the 2 wt% GO doped composite. This remarkable increase in T_g is observed as the filler loading increases beyond 1 and 2 wt%. The increase in T_g can be ascribed to an effective attachment of PVA to the GO sheets that constrains the segmental motion of PVA chains by hydrogen bonding and electrostatic attraction that prevents the segmental motions of the polymer chains.^20,21 This is the evidence of rheological percolation of the altered interphase within the polymer. Below 2 wt% doping, the vast interfacial area created by well-dispersed GO nanoflakes can affect the behaviour of the surrounding PVA matrix for several radii of gyration. This creates a co-continuous network of dramatically altered polymer chains and above 2 wt% this co-continuous network of polymer chains due to the above percolation behaviour of the polymer/GO matrix faces constrains in the segmental motion of PVA chains by hydrogen bonding and electrostatic attraction. This fundamentally changes T_g and beyond 2 wt% GO doping the T_g of the nanocomposite broadens and decreases.

The DSC thermograms show that the melting temperatures (T_m) of pure PVA and GO doped PVA composite films with 1, 2, 3, 4 wt% and 5 wt% GO doping are 193, 195, 197, 188, 198 and 203 °C respectively. The increase in T_m with the addition of GO confirms the interaction between PVA and GO. However, the increase in T_m of GO doped PVA indicates that not only hydrogen bonding but also additional effects are involved (e.g., steric effects) in the composites. In addition, beyond 2 wt% of GO, a sharp melting peak with the lower T_m and exothermic peaks with respect to PVA is observed. This indicates that the rest of GO (beyond 2 wt%) in the polymer composite behaves as inert material in the nanocomposite and the GO oxide platelets are responsible for the changes in the properties and these changes in the nanocomposites affect the mobility of the chains.

The thermal degradation studies of pure PVA and GO doped PVA composites were performed using thermogravimetric analysis (TGA) as shown in Fig. 4. The addition of GO improves the thermal stability of PVA and also the maximum mass loss temperature (T_max) is shifted toward higher temperatures and the degradation rate becomes slower, which indicates the involvement of a strong interaction between PVA and GO, leading to improvement of the thermo-stability of the composite films. Since, the thermal degradation of a polymer begins with chain cleavage and radical formation, the carbon surface of GO in the composite films might act as a radical scavenger to delay thermal degradation and hence improves the thermal stability of PVA.

Fig. 4 TGA thermograms of PVA and GO doped PVA composites (inset is TGA of GO).

The thermal degradation behavior of the PVA/GO composite consisting of an initial weight loss (∼8%) below 100 °C is attributed to the removal of moisture and water groups, as graphene oxide is known to store water in its stacked structure. Weight loss (∼18%) occurs at 100–270 °C, because of the removal of the functional groups such as hydroxyls, epoxides, and carboxylic acids on both the basal plane and periphery of graphene oxide, followed by a major weight loss (∼70%) between 270 °C and 430 °C, associated with the decomposition of oxygen-containing groups and the carbon backbone. This degradation of the polymer has mildly shifted to a higher temperature range with respect to the pure PVA matrix, which further confirms that the GO layers covered the surface of PVA well. The thermal stability of the nanocomposites was improved to a certain extent and here, the physical barrier effect of graphene oxide plays a vital role in the polymer composite. The functional unit of nanostructured GO forms a molecular chain with the hydroxyl-terminated PVA matrix, which restrains PVA from future degradation and this is considered as the main reason for the improved thermal stability.^2,22

Fig. 5 shows X-ray diffraction profiles of PVA, PVA/GO composites and GO. The diffraction peak of pure PVA appeared at 2θ = 19.68° resulting from its (101) crystal planes of PVA and the characteristic peak of GO sheets appears at 2θ = 9.32°, corresponding to the d-spacing of 0.948 nm. Such a value is much larger than that of natural graphite due to the generation of oxygen-containing functional groups between layers.

Fig. 5 WXRD of pure PVA, GO and GO doped PVA composites.

However, after GO was dispersed into the PVA matrix, the peaks at 2θ = 9.32° corresponding to GO almost disappear entirely in the XRD patterns, suggesting that GO was fully exfoliated into the polymer matrix and the appearance of some new diffraction peaks are seen, indicating clearly that GO was fully exfoliated into the polymer matrix, hence, the regular and periodic structure of GO is disappeared.²³ The trend of the peak shifts at 2θ = 18.9° of GO–PVA films and decrease of this peak with the addition of GO, might be attributed to the steric barrier effect of GO.²⁴ The absence of the graphitic peak at 2θ = 26.51° in the case of PVA/GO composites confirms better exfoliation of stacked layers into the PVA matrix and the new peaks appeared at 2θ = 13.4°, 16.2°, 24.8°, 27.8°, 31.2°, 35.8° through its noncovalent functionalization with the PVA polymer. In addition, the crystalline structure of the matrix was slightly affected due to the incorporation of GO.

Fig. 6 shows the typical stress–strain curves of the PVA film and PVA/GO composite films. Initially, the graphs exhibit an approximately linear regime from which the modulus can be calculated. The stress–strain curve shows that the PVA/GO composite exhibits maximum hydrogen-bond content during the elastic response. The PVA/GO composite structure reveals a linear response until rupture of the hydrogen-bonding network, followed by yielding from ∼3 to ∼5% strain.

Fig. 6 Stress–strain curve of pure and GO doped PVA blend films (inset stress–strain curve of PVA).

Pure PVA possesses a tensile strength of 3.3 MPa, a Young’s modulus of 7.6 MPa, stiffness of 378 N m⁻¹ and an elongation at break of 536.6%, when GO is incorporated into PVA; a dramatic increase in tensile strength, stiffness, and Young’s modulus of the composites is seen. The 1 wt% GO prepared composite showed an increase in tensile strength to 41.2 MPa and Young’s modulus to 1704 MPa. This enhancement of tensile strength and Young’s modulus of the PVA/GO composites could be attributed to strong hydrogen bonding between the higher density of oxygen-containing functional groups on the GO surface and surrounding molecules of PVA. The tensile strength of the PVA/GO composites increased from 41.3 to 87.9 MPa (47% increase) as the GO content increased from 1 wt% to 3 wt% and Young’s modulus also increased from 1704 MPa to 3390 MPa, as the GO content increased from 1 wt% to 3 wt%, attributed to the increase in the hydrogen bonding between PVA and GO composites.

In detail, the energy release during the formation of the hydrogen bond between PVA and GO is higher than the energy loss caused by the deformation of the PVA molecule. Thus, a high oxidation degree of GO, which could increase the probability of hydrogen bonds between oxygen-containing groups (e.g., hydroxyls, carboxyls and epoxides) of GO and the hydroxyl groups of the PVA chains. This would enhance the strength of hydrogen bond networks occurring between the GO sheet and PVA matrix; this interaction provides better load transfer, hence the increase in the mechanical properties.^25,26

The incorporation of GO sheets simultaneously improved the strength and elongation at break up to 3 wt% attributed to the good interaction with the PVA matrix and slipping of PVA molecules at the surface of GO in the tensile testing.²⁵ Loading more than 3 wt% of GO in PVA/GO composites shows a decrease in the elongation at break, indicating that the rigidity of PVA was increased because of the restriction on free movement or mobility of the PVA segments during extension. The increase in the stiffness with GO content in PVA/GO composites is attributed to the presence of cooperative interactions between non uniformly distributed small clusters of hydrogen bonding groups located in the inter-sheet gallery. Such cooperativity enhances the stiffness of the nanocomposites. It was observed that the tensile strength and Young’s modulus decreased when the loading was more than 3 wt% in PVA/GO composites, attributed to the agglomeration of GO which could reduce the effective cross-linking points and interaction between PVA and the GO polymer matrix.^27,28

The dc conductivity of GO doped PVA composites are studied from 303 K to 343 K and shown in Fig. 7. The conductivity of pure PVA and GO doped nanocomposites, were measured by a four probe method. The pure PVA shows low conductivity (∼2.6 × 10⁻¹⁰ S cm⁻¹), however, the increase of 4 orders of magnitude for 1 wt% GO doped PVA (5.71 × 10⁻⁶ S cm⁻¹), was several magnitudes higher than that of pure GO. Although GO was nonconductive, the conductivity in PVA/GO nanocomposites was boosted due to π–π stacking between GO and PVA. The increased conductivity can be assigned to the percolation behaviour of GO. That is, conductivity of the polymer composites remains at low values until a critical loading content is achieved. Then, it increases by several orders of magnitude and for our prepared GO doped PVA samples the percolation threshold lies between 1 and 2 wt% of GO.

Fig. 7 Variation of log(σ_dc) with temperature of GO doped PVA composites.

The higher percolation threshold in our case seems reasonable because of the presence of non-exfoliated graphite, which diminishes the filler aspect ratio. The low current for GO doped PVA composites may be related to the thickness of the polymer film (100–150 µm), particularly, the ratio between the quantity of the polymer powder and water; the used amount of water for dissolving the GO and PVA powder could have led to low concentrations of the conducting material in the obtained PVA/GO solution. As a result, when the solution was dried, numerous transparent (probably non-conductive) areas appeared.

In pure PVA, since there is no possibility of charge injection from the electrodes and the origin of the measured current for pure PVA is attributed to the motion of ions in the polymer matrix, originating from impurities, these impurity ions can travel through the whole sample between two electrodes; the increase in the temperature facilitates impurity ions toward the electrode, hence the increase in the conductivity of PVA. For 1 wt% and 2 wt% GO samples, the conductivity increased several orders of magnitude, due to the intrinsic field-dependent conductivity of GO. The abundant surface groups present in GO lead to a disrupted sp² structure and act as energy barriers for charge transport along the carbon network in the polymer matrix. These blocked electrons slowly become capable of tunnelling through the GO with the electric field and temperature. Thus the electronic conduction along the GO network is facilitated and becomes the major contribution to the total current in the polymer samples. As GO was fully exfoliated into the polymer matrix and for a higher weight% of GO, GO may not find a path toward electrodes or will be completely constrained in narrow spaces inside the polymer network. Thus the current drops as impurity ions encounter GO, and the conductivity of the composites is reduced.^14,15

Fig. 8(i) and (ii) shows the variations of dielectric constant (ε′) and dielectric loss (ε″) of PVA/GO composite films with a frequency range of 4 Hz to 5 MHz at room temperature. At low frequencies, the dielectric constant attained higher values because complete orientation of the molecule in the polymer matrix. At higher frequencies, the sharp decrease in dielectric constant indicates that the polarization can no longer follow rapid changes in the external electric field and relaxation can occur and the dielectric constant increases in the lower frequency region and shows that the maximum for the 2 wt% GO doped composites is due to the local motion of the side group dipoles about the main backbone of PVA.²⁹

Fig. 8 Variation of frequency dependent (i) dielectric constant (ε′) (ii) dielectric loss (ε″) of pure and GO doped PVA composites.

The dielectric loss at the low frequency region becomes very high due to free motion of dipoles within the polymer composites and an increase in the dielectric loss with the increase of GO content is attributed to the increase in polar groups. The increase in the dielectric constant and dielectric loss with GO up to 2 wt% is ascribed to the motion of free charge carriers due to the formation of a continuous conductive pathway between and throughout the medium. According to the Maxwell–Wagner–Sillars (MWS) process polymer–filler interfacial (like donor–acceptor complexes) is necessary to get changes in dielectric properties.

Fig. 9(i) shows the plot of the dielectric constant at different temperatures for 2 wt% GO doped PVA composites. The dielectric constant increases with the increase in temperature and the value of dielectric constant decreases with increasing frequency up to 1 MHz at all the temperatures; the decrease in the dielectric constant with an increase in frequency is attributed to the increase in interfacial polarization with the incorporation of GO. Above 1 MHz frequency, the dielectric constant remains almost constant for different temperatures.

Fig. 9 The plot of the (i) dielectric constant (ε′) (ii) dielectric loss (ε″) at different temperatures for 2 wt% GO doped PVA composites.

Fig. 9(ii) shows the variation of dielectric loss with frequency at different temperatures; the value of dielectric loss increased with frequency. At lower frequency, the increase in dielectric loss is attributed to the increase in segmental mobility of polymer chains and the relaxation peak of the PVA/GO (2 wt%) composite was shifted towards a higher frequency region showing that the relaxation process increases with temperature. This could be because, as the temperature increased, the polar group present in the polymer matrix will become more free to orient, allowing it to keep up with the changing electric field and an increase in the micro-Brownian motion of the polymer chain takes place, hence a shift toward higher frequency could be expected.³⁰

According to the percolation theory, the dielectric and conductive properties of composites change sharply near the percolation threshold. To reach the percolation threshold, PVA films containing different contents of GO were studied. Graphene oxide flakes can provide percolated pathways for electron transfer making the composites electrically conductive. As the frequency increased, the particle orientation also plays an important role: the percolation threshold becomes greater as particles are aligned parallel. The percolation theory depicts that the variations of ac conductivity with frequency follows a power law as the filler content approaches percolation threshold as shown in Fig. 10(i) and highest conductivity observed for 2 wt% of GO doped PVA is 5.74 × 10⁻⁵ S m⁻¹ in the higher frequency region. The temperature dependence of ac conductivity (as shown in Fig. 10(b)) for a 2 wt% GO doped PVA nanocomposite within the temperature range of 303 and 343 K is seen. As temperature increases, the conductivity value of GO doped nanocomposites also increase; this is due to the increase of thermal movement of polymer chain segments.

Fig. 10 Variation of frequency dependent (i) ac conductivity (σ_ac) at room temperature and (ii) ac conductivity (σ_ac) of 2 wt% GO doped PVA composites at different temperatures.

Fig. 11(i) and (ii) shows the pressure dependent dielectric constant and loss of 2 wt% GO doped PVA composites under different applied pressures at a temperature of 303 K. The applied pressure was longer than 20 min for a composite and testing was repeated several times. The dielectric constant increases with increasing frequency at all applied pressures. As it is known that the dielectric constant and dielectric loss is dependent on polarisability and free volume of the constituent element present in the composites, this free volume size of the constituent decreases when the pressure is applied on the composites. An increase in the dielectric constant and loss under different applied pressures might be expected and the consequent intensification of the atomic forces. A positive response with the increase in the applied pressure may be due to the inter-chain relaxation process occurred in the polymer composites and very small proportion values in the dielectric constant and loss is due to the presence of free water molecules in the system.^31,32

Fig. 11 Variation of the (i) dielectric constant (ε′) (ii) dielectric loss (ε″) for 2 wt% of GO doped PVA composites at different pressures.

In Fig. 12, as frequency increased, an increase in the dielectric constant and dielectric loss factors is seen, which reflects a stronger molecular interaction occurring between GO and PVA. Hence, the dielectric constant and dielectric loss overlaps each other and frequency dependent studies of the 2 wt% GO doped composite shows an increase in the conductivity with the increase in the applied pressure. At an atmospheric pressure of 0.48 atm, the drastic increase in the conductivity that is seen can be attributed to a decrease in the free volume size of the constituent. With a further increase in the applied pressure there is no such enhancement in the conductivity. This can be attributed to minimal variations of interaction constants and changes in the domain structure.^13,15,33

Fig. 12 Tunable ac conductivity properties of the 2 wt% GO doped PVA with different applied pressures and experimental set up for applied pressure studies.

Nyquist plots of the all-solid-state composites with different GO-doped PVA and PVA composites are as shown in Fig. 13(i) and were obtained by a frequency response analysis of the frequency range from 4 Hz to 5 MHz. The Nyquist plots exhibit an intercept on the real axis (x axis) at high frequency (close to 5 MHz); a plot of the imaginary component of the impedance against the real component shows a decrease in the semicircle, which is related to the lowest charge transfer resistance, and a transition to linearity at low frequency which exhibits an ideal capacitive behaviour. The complex impedance plots showed two well-defined regions: a semicircle in the high frequency range related to the conduction process in the bulk of the complex and a linear region in the low frequency range, which is attributed to the bulk effect of blocking electrodes. Compared to pure PVA films, GO doped PVA films have lower ionic resistance and, as mentioned, highest ionic conductivity is observed for 2 wt% GO, this indicates the existence of a maximum and effective interaction between DO functional units and the PVA polymer matrix in the nanocomposite. The temperature dependent Nyquist plot for the 2 wt% GO doped PVA nanocomposite sample as a function of temperature is shown in Fig. 13(ii). As observed, the ionic resistance of the films decreases with an increase in temperature; this can be attributed to the increase in segmental motion of GO doped PVA nanocomposites, which hence shows an increase in ionic conductivity.²⁹

Fig. 13 Nyquist plots of GO doped PVA doped nanocomposite (i) at room temperature and (ii) at different temperatures of 2 wt% GO doped nanocomposites.

The section regarding GO doped PVA will summarise the observations that can be made using TEM imaging and spectroscopic techniques, such as determination of the number of graphene layers, detection of atomic scale defects (vacancies, dislocations, grain boundaries, etc.), and chemical characterisation. Fig. 14 shows an atomic resolution TEM micrograph acquired under imaging conditions where GO clearly illustrates flake-like shapes of a few hundred nanometers in size staked together. GO is prepared by the modified Hummer’s method and TEM images of the synthesized graphene oxide illustrate the flake-like shapes of graphene oxide and conversely it was observed that GO single sheets were homogeneously dispersed within the polymer matrix with hardly any aggregation. This observation is well supported by the XRD study which showed good dispersion of GO in the polymer composites.³⁴

Fig. 14 TEM image of graphene oxide.

Atomic force microscopy (AFM) has emerged as an invaluable tool for the characterisation of GO in terms of determining the shape of the deposited flakes and the level of increase in the flake and also AFM allows for 3D profiling of surfaces at the nanoscale. The topography of GO doped PVA is as shown in Fig. 15.

Fig. 15 AFM images of graphene oxide doped PVA.

It is well known that graphene oxide has a variety of functional groups such as epoxides (bridging oxygen atoms), carbonyls (=CO), hydroxyls (–OH) and phenol groups attached to its surfaces. The attachment of these groups with the polymer leads to sp³ hybridisation which leads to a certain amount of buckling of GO in the polymer composites. Thus, the interlayer spacing between GO sheets is larger than in graphite and also it might be expected that the height measurements obtained by atomic force microscopy will exceed that of graphene. The exfoliated graphene oxide showed an average GO flake size of an individual of 200–300 nm and some high regions showing an average height of 10 nm. The bumpy surface is attributed to regions of dead space because of extensive edge functionalization.³⁵

4. Conclusions

A novel GO based PVA polymer nanocomposite has been prepared by the solution casting technique. The FTIR study confirms the formation of intramolecular and intermolecular hydrogen bonding with the addition of GO. The higher intensity and larger width of this band in the GO doped PVA spectrum suggests the formation of CTC with OH groups of the PVA matrix. The UV visible absorption spectrum showed an increase in the absorption of the GO doped composite. This is due to an increase in the hydrogen bonding with PVA and in the XRD result, the peaks at 2θ = 9.32° corresponding to GO almost disappear suggesting exfoliation of GO into the polymer matrix and the appearance of new peaks clearly demonstrate that GO was fully exfoliated into the polymer matrix. In the case of mechanical studies, 3 wt% of GO doped PVA showed a maximum tensile strength of 86 MPa and Young’s modulus of 3390 MPa. The dielectric measurements of GO doped PVA composites at room temperature obeyed the Maxwell–Wagner–Sillars (MWS) effect and it is found that the maximum dielectric constant (ε′) and loss (ε″) is observed for 2 wt% composites. The temperature dependent dielectric constant and loss increases at all the temperatures attributed to the increase in interfacial polarization with the increase in temperature. The TEM/AFM micrograph of GO clearly illustrates the flake-like shapes of few hundred nanometers in size staked together. It is found that the pressure dependent dielectric properties of GO doped PVA confirmed that these composites have excellent potential to revolutionize the use of nanocomposites and enable their widespread use in large-scale applications. Hence, the prepared composites provide an exciting opportunity for pressure sensing applications.

Acknowledgements

The authors are thankful to Department of Atomic Energy, Board of Research in Nuclear Sciences (BRNS), Govt. of India for the research project (2010/37C/7/BRNS/832) and Department of Science and Technology (DST), Govt. of India for the research project (SR/FTP/PS-011/2010) and (SB/EMEQ-089/2013) and also the authors are thankful to USIC, Karnatak University, Dharwad for providing the AFM facility.

References

1. H. M. Kim, J. K. Lee and H. S. Lee, Thin Solid Films, 2011, 519, 7766.
2. D. Hana, L. Yana, W. Chena, W. Li and P. R. Bangal, Carbohydr. Polym., 2011, 83, 966.
3. Y. Pan, T. Wu, H. Bao and L. Li, Carbohydr. Polym., 2011, 83, 1908.
4. P. K. Pramoda, Y. K. Mya, T. T. Lin, X. Lu and C. He, Polym. Eng. Sci., 2012, 2530.
5. Y. Wang, Z. Shi and J. K. Yin, Polymer, 2011, 52, 3661.
6. C. O. Compton, W. S. Cranford, W. K. Putz, Z. An, C. L. Brinson and J. M. Buehler, et al., ACS Nano, 2012, 6, 2008.
7. S. Morimune, T. Nishino and T. Goto, ACS Appl. Mater. Interfaces, 2012, 4, 3596.
8. A. Kundu, K. R. Layek, A. Kuila and K. A. Nandi, ACS Appl. Mater. Interfaces, 2012, 4, 5576.
9. S. Stankovich, A. Dikin, H. Dommett, M. Kohlhaas, J. Zimney, A. Stach, D. Piner, T. Nguyen and S. Ruoff, Nature, 2006, 442, 282.
10. T. Ramanathan, A. Abdala, S. Stankovich, D. Dikin, M. Herrera-Alonso, R. Piner, D. Adamson, H. Schniepp, X. Chen, R. Ruoff, S. Nguyen, I. Aksay, R. Prud’hommem and L. Brinson, Nat. Nanotechnol., 2008, 3, 327.
11. I. Sakellis, A. Papathanassiou and J. Grammatikakis, Synth. Met., 2010, 160, 2228.
12. M. Naoki, K. Takuya and N. Yuta, Sci. Rep., 2016, 6, 21715.
13. S. M. Imran, G. N. Shao, M. S. Haider, H. J. Hwang, Y. Choa, M. Hussain and H. T. Kim, Polym. Eng. Sci., 2016, 56, 1031.
14. S. Baradaran, E. Moghaddam, J. W. Basirun, M. Mehrali, M. Sookhakian and M. Hamdi, et al., Carbon, 2014, 69, 32.
15. S. G. Rathod, R. F. Bhajantri, V. Ravindrachary, T. Sheela, P. K. Pujari, J. Naik and B. Poojary, J. Polym. Res., 2015, 22, 6.
16. N. Ashok Kumar, J. H. Choi, R. Y. Shin, W. D. Chang, L. Dai and B. J. Baek, ACS Nano, 2012, 6, 1715.
17. T. Zhang, P. Liu, C. Sheng, Y. Duan and A. Zhang, Chem. Commun., 2014, 50, 2889.
18. V. Saini, O. Abdulrazzaq, S. Bourdo, E. Dervishi, A. Petre and G. V. Bairi, et al., J. Appl. Phys., 2012, 112, 054327.
19. A. Mathkar, D. Tozier, P. Cox, P. Ong, C. Galande and K. Balakrishnan, et al., J. Phys. Chem. Lett., 2012, 3, 986.
20. Y. Li, T. Yang, T. Yu, L. Zheng and K. Liao, J. Mater. Chem., 2011, 21, 10844.
21. H. Feng, Y. Li and L. Li, RSC Adv., 2012, 2, 6988.
22. S. M. Parka, B. Ramaraj and R. K. Yoona, Appl. Surf. Sci., 2012, 258, 8656.
23. Y. Fenga, X. Zhanga, Y. Shena, K. Yoshino and W. Fenga, Carbohydr. Polym., 2012, 87, 644.
24. H. J. Yang and D. Y. Lee, Highly electrically conductive rGO/PVA composites with a network dispersive nanostructure, J. Mater. Chem., 2012, 22, 8512.
25. N. Ding, X. Chen, L. M. Wu and X. Lu, J. Phys. Chem. C, 2012, 116, 22532.
26. V. N. Medhekar, ACS Nano, 2010, 4, 2300.
27. C. O. Compton, W. S. Cranford, W. K. Putz, Z. An, C. L. Brinson, J. M. Buehler and T. S. Nguyen, ACS Nano, 2012, 6, 2008.
28. Y. He, N. Zhang, Q. Gong, H. Qiu, W. Wang and Y. Liu, et al., Carbohydr. Polym., 2012, 88, 1100.
29. J. Y. Kim, W. H. Lee, J. Y. Suk, J. R. Potts, H. Chou and I. N. Kholmanov, et al., Adv. Mater., 2013, 25, 2308.
30. A. Patra and D. Bisoyi, J. Mater. Sci., 2010, 45, 5742.
31. X. Li, J. Rong and B. Wei, ACS Nano, 2010, 10, 6039.
32. C. Fancgao, G. A. Saunders, E. F. Lambson, R. N. Hampton, C. Carini, C. Di Marco and M. Lanza, J. Polym. Sci., Part B: Polym. Phys., 1996, 34, 425.
33. X. Yan, J. Chen, J. Yang, Q. Xue and P. Miele, ACS Appl. Mater. Interfaces, 2010, 2, 2521.
34. Y. Pan, T. Wu, H. Bao and L. Li, Carbohydr. Polym., 2011, 83, 1908.
35. D. Chen, X. Wang, T. Liu, X. Wang and J. Li, ACS Appl. Mater. Interfaces, 2010, 2, 2005.

---