Flexible Conductive Nanocomposites for Electrochemical Devices Based on Chlorinated Natural Rubber/Nickel Oxide Nanoparticles

A study on structural, thermal, temperature-dependent electrical properties such as AC conductivity and dielectric properties of flexible conductive chlorinated natural rubber (Cl-NR) were carried out with various contents of nickel oxide (NiO) nanoparticles. The role of fillers on DC conductivity of the composites was correlated with different theoretical models. The FT-IR spectra showed the characteristic absorption band for nano-NiO in the rubber indicating the effective incorporation of nanoparticles in Cl-NR. Optical bandgap energy was observed minimum for 5 phr loaded composite. TGA results showed that the thermal stability increased with NiO content in the polymer matrix. The dielectric properties and AC electrical conductivity increased significantly with the temperatures and also with the addition of nanoparticles up to 5 phr loading. The activation energy of the electrical conductivity decreases with an increase in temperature for all the systems. The higher value of dielectric permittivity explained the electrode polarisations at the low-frequency region. The skewed semi-circular arc in the Cole–Cole plot observed for all the samples explained the semiconducting behaviour of the nanocomposites. Space charge polarisation and relaxation dynamics of Cl-NR composites were explained based on modulus spectra. The McCullough model was found to be the most promising one to explain the DC conductivity of the system which emphasizes the importance of interfacial interaction at the boundary of filler and the rubber chain for the network formation. According to the results of this study, these samples can be used in highly durable flexible electronic devices such as conductive sensors, actuators and super-capacitors.


Introduction
Most of the recent electronic technologies are based on a type of promising artificial elastomer muscles applied in the field of dielectric elastomeric sensors, actuators and transducers. These dielectric elastomer materials enable inherent compliance, light-weight, compactness, large actuation strain and high energy efficiency [1,2]. The uniqueness of dielectric elastomer materials includes its noiseless operations, high-speed activity, very low energy consumption, large deforming ability, as well as very high energy efficiency, flexible nature and economic viability [3]. The efficiency of these materials makes them an appealing substitute in the fields of end effectors in robotics and conventional actuators [4]. The advanced uses of potential electrostatic actuators and other dielectric elastomer materials include biomimetic robots, soft prosthetic devices, and compliant manipulators or grippers for delicate tasks such as surgeries [5,6]. Dielectric film sandwiched between docile electrodes comprises the dielectric elastomer material [7]. The combination of electrodes with dielectric elastomer film results in a versatile capacitor with a capacitance directly proportional to the relative permittivity of the material and area of the electrode surface, inversely proportional to thickness [8]. Accumulation of opposite charges on electrodes on the application of external voltage and the resulting coulombic force causes the reversible compression and expansion of the impeded dielectric elastomer material [9].
Materials having higher flexibility with good electrical and thermal properties are required for the applications of sensors, microwave absorbents, non-linear optics, braille displays, pumps, valves, grippers and micro positioning items [10,11]. Common dielectric elastomers include different acrylic polymers, polyurethane, silicone and natural rubbers [12,13]. Among them, natural rubber is highly flexible and the mouldability of NR gives a high degree of freedom to fabricate technologically useful products. Jamal et al. reported the magnetic and dielectric properties of nickel incorporated NR [14]. However, the electrical properties of NR composites are very poor due to the absence of reactive double bonds in the backbone isoprene units. The unsaturation present in the NR is a potential candidate for chemical modification to introduce polar groups in non-polar NR. Chlorination of NR remains the best choice in this context because it can result in a high-performance dielectric material with more polarizable groups.
During the past several decades the scientific community has stuck on the search for elastomer nanocomposites as they are technologically and commercially important [15,16]. The effective embedding of different nanofillers in elastomer matrix attracted scientific interest as they markedly enhance the dielectric, electrical and sorption characteristics of elastomers [17][18][19]. However, the metal oxide nanoparticles have a major role in imparting electrical conductivity, dielectric behaviour and thermal stability to elastomeric systems. Among the metal oxide nanoparticles, nickel oxide particles are chemically more stable with high dielectric properties and therefore make a better choice for enhancing the electrical properties of insulating polymers [20][21][22].
In recent years, demand for multifunctional flexible nanocomposite meeting the demand for electronic devices has been greatly increased due to significant expansion in the field of electrically active polymer composite processing [23,24]. Works on the utilization of nickel oxide nanoparticles as a potential filler to improve the optical and thermal stability of chlorinated rubber and studies on the effect of temperature on electrical conductivity and dielectric properties of chlorinated natural rubber nanocomposites are not yet reported. Therefore, the purpose of this study is to investigate the effect of temperature on AC conductivity, different dielectric parameters such as dielectric constant, dielectric loss, dielectric modulus of Cl-NR/NiO nanocomposite with various concentrations of metal oxide nanoparticles at different frequencies. The enhancement in DC conductivity with the addition of different contents of NiO to Cl-NR has been compared with various theoretical modelling to establish the mechanism behind the electrical conductivity of the prepared composites. Also, the effect of metal oxide nanoparticles on the thermal stability and optical properties of chlorinated rubber has been examined with respect to the different loading of nanofillers.

Preparation of Chlorinated NR/Nickel Oxide Nanocomposites
The chlorinated NR with different contents of NiO nanocomposites were prepared by a two-roll mill mixing technique at room temperature. Firstly, 100 g of Cl-NR was masticated in a two-roll mill at a frictional ratio of 1:1.25. Different loadings (0, 1, 3, 5, 7 and 10 phr (parts per hundred parts of rubber)) of NiO nano-powders caped with CTAB were added to the masticated Cl-NR followed by mixing with 6 phr DCP (curing agent) and continued mixing for 8 min as per the ASTM D-15-627. The nanocomposites were cured at 150 °C in an electrically heated press up to the optimum cure time obtained from the rubber processing analyzer. The optimum cure time obtained for 0,1, 3, 5, 7 and 10 phr samples were 14.4, 13.3, 12.2, 11.6, 10.4 and 9.0 min respectively. The vulcanized sheets with a thickness of 2 mm were used for the polymer testing. The schematic representation for the fabrication of Cl-NR/NiO nanocomposites is illustrated in Scheme 1.

Characterisations
Structural elucidation of Cl-NR/NiO nanocomposites was analysed using a Fourier transform infrared spectrometer (FT-IR series model JASCO 4100). The unvulcanised samples were dissolved in toluene and the solution was placed on the surface of KBr pellets in the scanning rate of 20 with a resolution of 4 cm −1 with the wavenumbers between 4000 and 400 cm −1 . The ultraviolet-visible (UV-Vis) absorption spectra of Cl-NR and different contents of NiO nanoparticles inserted Cl-NR composites in toluene were recorded in a Shimadzu-2600 UV-Vis spectrometer. The thermal stability of the composites was evaluated by a Hitachi STA7200 thermogravimetric analyzer with pure nitrogen with a flow rate of 20 ml/min and at a heating rate of 10 °C/min. Circular disc samples were punched out from the vulcanized sheet for electrical property studies. AC conductivity, dielectric properties such as dielectric constant, dielectric loss and dielectric modulus of Cl-NR/NiO composites were determined with the Hioki impedance analyser (HIOKI 3570 model) in the frequency range of 10 2 to 10 6 Hz by varying temperature from 30 to 90 °C. The dielectric permittivity of the composites was calculated by using the relation, and dielectric loss by the equation where ε′ is the dielectric permittivity of the material, C is the capacitance, 0 is the permittivity of free space, A is the area of cross-section of the material, δ is the loss tangent and d is the thickness of the sample.

Fourier Transform Infrared Spectroscopy (FT-IR)
The FT-IR spectra of NiO filled Cl-NR composites are depicted in Fig. 1. The stretching vibrations of CH, CH 2 and CH 3 groups of natural rubber are present at 2862 cm −1 , 2928 cm −1 and 2963 cm −1 respectively [26]. The absorption peak at 1381 cm −1 corresponds to the methyl group and the band at 981 cm −1 is the cis CH=CH of NR [27]. The wagging band of =CH in NR appeared at 833 cm −1 is the typical fingerprint region of the rubber. Characteristic C-Cl absorption peak observed at 775 cm −1 along with the presence of cyclopropyl segments at 1094 cm −1 indicates the stereospecific addition of dichlorocarbene unit to NR [28].
There is a shift of 1456 cm −1 corresponding to the C-H bending vibration in pure natural rubber to 1448 cm −1 is due to the steric hindrance of the cyclopropyl ring present in the Cl-NR. It can be observed from the figure that the penetration of chlorine to the main chain of polyisoprene reduces the intensity of the olefinic double bond at 1664 cm −1 . Characteristic shifts in peak position due to interaction of new chemical group and the formation of new bonds along with the inherent bonds in polyisoprene can also be observed. The FTIR spectra of NiO showed the absorption band at 3437 and 1632 cm −1 are attributed to the hydroxyl group on the surface of NiO and the strong peak at 501 cm −1 is assigned to the typical Ni-O bond [29]. The spectrum of the nanocomposite exhibits most of the vibrations of chlorinated rubber with the strong band of NiO at 511 cm −1 . The OH vibration of NiO (3437 cm −1 ) in the composite is shifted to a higher wavenumber of around 3452 cm −1 . The shift of peaks in the nanocomposites with the presence of NiO indicating the effective incorporation of nanoparticles in the polymer matrix.

Ultra Violet-Visible Spectra (UV-Vis)
The optical properties of chlorinated NR with different loadings of nano NiO are shown in Fig. 2. The UV absorption spectra of chlorinated NR show a weak band at 290 nm corresponding to π-π* transition of π bonded electrons in the backbone isoprene chain. Yu et al. reported that the UV-Vis spectra of chlorinated natural rubber degraded at 160 °C showed the absorbance at 245 nm is due to the π-π* transition of conjugated polyene present in NR [30]. The composite with 1, 3, 5, 7 and 10 phr of NiO exhibit the UV peaks at 292, 294, 296, 296.5 and 298 nm respectively. It is clear These changes observed in UV spectra are maximum in 5 phr loaded sample, explain the strong interaction of nano NiO with the chlorinated matrix. The intensity of the UV peak of composite decreases beyond 5 phr loading indicating the formation of nanoparticle agglomerates. The greater stress developed in the chlorinated matrix by the insertion of nanofiller is due to the short distance between the adjacent particles at higher loading of fillers.

Optical Bandgap-Tauc Plot
Tauc's relation is used to determine the optical bandgap of the composites and is plotted in terms of (αhν) 2 and hν as depicted in Fig. 3.
The Tauc's relation is associated with the photon energy (hν) and optical bandgap energy (Eg) is given in the above expression (Eq. 3), where B is an arbitrary constant and 'n' is an index indicating the direct (n = 0.5) or indirect nature (n = 2) of bandgaps [31]. The optical bandgap energy of Cl-NR with different loading of NiO is presented in in Fig. 4. It is clear that the optical bandgap of composites is found to decline continuously with increased filler loading up to an equilibrium loading level (5 phr) and then slightly increases. The regular change in the bandgap energy is associated with the structural change of Cl-NR with increased filler content [32]. The movement of electrons from the valence band to the conduction band is facilitated by the generation of new energy states observed in the composites. The strong interaction between NiO and the polymer is maximum at 5 phr  Fig. 2) which alters the Fermi level thereby replacing the valence band nearer to the conduction band, which could narrow the distance between CB and VB. [33]. Pradeep et al. reported a similar trend in optical bandgap energy for starch incorporated natural rubber [34]. It can be concluded that the increased loading of NiO nanofiller lowers the bandgap energy.

Thermogravimetric Analysis (TGA)
Thermogravimetric analysis for the thermal stability behaviour of fabricated composites is detailed in Fig. 5. It is observed that all the samples show a two-stage degradation behaviour [35]. Chlorinated NR shows an initial thermal degradation at a temperature of 162 °C corresponding to non-rubber impurities present in the chlorinated matrix, vulcanising agent, or removal of hydrogen chloride [36]. The final and major degradation of about 80% is found at 316 °C corresponding to the rupture of the main chain of the chlorinated NR. Zhong et al. reported the two-stage thermal oxidative decomposition of chlorinated natural rubber and this is in good agreement with our study [36]. The curves for the composites with 5 and 10 phr filler loading show the initial minor degradation at a slightly higher temperature such as 169 °C and 173 °C respectively. The major degradation of composites is observed at 321 °C and 328 °C respectively for 5 and 10 phr filler loading. This hike in the degradation temperature is observed in composites due to the good dispersion of nano nickel oxide and the reinforcement provided by the filler-polymer interfacial adhesion. Chen et al. reported a similar trend in degradation temperature for poly (lactic acid)/liquid natural rubber/nickel-zinc ferrite blend nanocomposites [37]. The uniform dispersion of filler particles forms an effective surface coating on the composites and thereby protects the surface from further degradation causing a high amount of char residue compared to chlorinated NR. This higher char residue indicates the induced flame resistance in composite materials by the effective filling of metal oxide nanofiller particles. This result is in good agreement with NBR/hydroxyapatite nanocomposite [2]. This detailed thermal degradation analysis is useful in understanding the high-performance material for hightemperature applications.

Temperature Dependent AC Conductivity
AC conductivity plots of Cl-NR and its composites of varying combinations at different temperature regions (30-90 °C) are illustrated in Fig. 6. The special nature of logarithmic plots of AC conductivity and frequency explains the semiconductor nature of the parent Cl-NR and the increased conductivity values indicate the accelerated conductivity provided by the metal oxide filler particles [38]. It is observed from the figure that the conductivity increases with applied temperature and can be attributed to the accelerated mobility of charge carriers. Hopping type of conduction mechanism is possible in these systems as observed from the plots of the logarithm of AC conductivity vs frequency. The possible polarisation effect is the main criteria behind the increase in conductivity at higher frequencies, along with this hopping of charge carriers also contributing to higher conductivity possible at these higher frequency regions. A similar trend in AC conductivity and the activation energy was reported by Hashim et al. in their investigation on the influence of manganese zinc ferrite in rubber matrix [39]. They explained the hopping of charge carriers between Fe 2+ and Fe 3+ ions, the Maxwell-Wagner two-layer model exhibited by ferrites is used to explain the conductivity behaviour in the rubber-ferrite systems [39]. The presence of nickel oxide nanofiller effectively enhanced the electrical conductivity of Cl-NR and the maximum conductivity among the composites is obtained for 5 phr loaded composite. The possible charge hopping and interfacial polarisation effects are more pronounced in composites due to the effective incorporation of nanofillers and thereby the increased number of charge carriers. The loading of filler affects the charge hopping as the distance between the charges decreases with the increase in filler loading and the hopping process becomes more facile [20]. The poor conductivity observed for 7 and 10 phr composites can be explained on the basis of the clustering of filler particles beyond the equilibrium filler concentration (5 phr). These clusters of fillers separate the charge carriers through a large distance and therefore the hopping rate is diminished. The slope of these plots of AC vs frequency can be utilized to calculate the parameter called exponential factor present in the Power law equation (σ AC = Aω s ). Exponential factor value (s) usually falls between 0 and 1 [40]. The nature of conduction and the extent of interaction of charge carriers present in polymer lattice can be explained on the basis of this s value. There are two common theories possible for the explanation of the conduction mechanism of charge carriers, known as small and large polaron assisted tunnelling [41]. The positive variation of s value with temperature indicates the large polaron mechanism and similarly, the negative influence of temperature on s value is based on small polaron theory. The 's' value for the composites is found to be decreased with temperature (Table 1) indicating the small polaron conductivity mechanism. The activation energy can be calculated from the plot of 1000/T vs log AC conductivity (Fig. 7), it is observed that the activation energy is very small in the case of composites (Table 2). This small value of the energy of activation is a clear indication of the high hopping conduction by the electronic conduction mechanism rather than ionic conduction [41].

Dielectric Analysis
The mechanism of ion transportation and phase transitions in polymer composites can be analysed in a simple way using dielectric analysis [42]. Figure 8 shows the variation of dielectric constant and dielectric loss tangent with   frequency for chlorinated NR and 5 phr composite at different temperatures. The higher dielectric values confirm the electrode polarisation occurred at lower frequency regions and this nature explains the non-Debye type of conduction mechanism possible in the above systems [43]. As the frequency increases the decrease in dielectric constant becomes rapid and finally, the dielectric constant is found to become independent of frequency. Similar behaviour of dielectric permittivity and dielectric loss with frequency is observed for multiwalled carbon nanotube/rubber nanocomposites [44]. The decreasing trend of dielectric permittivity with increasing frequency is observed for carboxylated nitrile rubber nanocomposites as reported by Manna and Srivastava [45]. This independent nature can be attributed to the inability of charge carriers or dipoles in polymer composite to translate and orient in the direction of the applied external field. At high frequencies, there is no excess ion diffusion in the applied field direction as the periodic reversal of field occurs very rapidly. In other words, at higher frequencies the interfacial electrical dipoles possess only less time interval to orient in the field direction of alternating current, therefore the charges cannot follow the field and the contribution to permittivity ceases. The value of dielectric constant and dielectric loss values decreases with frequency as the polarisation due to charge accumulation decreases at higher frequencies. This relation between dielectric permittivity and frequency is related to Maxwell-Wagner type dispersion caused by the presence of an interfacial surface [46]. The host polymer usually possesses a lower dielectric constant than the composites despite the presence of ion pairs during the dielectric measurement. The temperaturedependent plots also show the same trend as that at room temperature conditions. Low dielectric constant value at a higher frequency and high value at low frequency edge due to the electrode polarisation process. The presence of localised charges at the interface of electrodes causes this phenomenon of polarisation. This polarisation phenomenon can result from the presence of components called the polymer matrix or the bare polymer and the metal oxide nanofillers with different conductivities when coming in contact with each other causing the electrical double layer and resulting in an interfacial polarisation effect. In the case of dielectric loss, it is caused by the polarisation which is in phase with the applied alternating electric field. Space charge polarisation at low frequencies causes high dielectric loss and at high frequencies, dielectric loss value decreases similar to dielectric constant due to the increased electrical conductivity of samples. From the figures of dielectric permittivity can also be concluded that as the temperature increases both dielectric constant and dielectric loss values increases.

AC Impedance Analysis
The dependence of AC impedance on frequency for Cl-NR and its composites is plotted in Fig. 9 at various temperatures. The real part of impedance consists of both frequencydependent and independent regions in their spectra. The impedance of bare polymer is high as the conductive network of electrons or charge carriers is low. But in the case of composites as the filler loading increases the NiO particles come closer and form a large network of conducting electrons, thereby the conductivity increases and the impedance is found to be decreased appreciably. After attaining a critical frequency, the hopping of charge carriers becomes more feasible as they are sufficiently excited and therefore the impedance reaches a constant value indicating the frequency-independent nature. The imaginary counterpart of impedance is exhibited in Fig. 9b and d. The peak found in the frequency region indicates the dipolar relaxation possible in the systems, the values of imaginary impedance increase with applied frequency and reached a maximum value ( Z II max ) and depreciation is found thereafter [47]. An appreciable decrease in impedance value is found in the imaginary part also, explaining the comprehensive depreciation of Z II resulting from the hike in conductivity provided by the metal oxide fillers. The peak maximum corresponding to the relaxation is shifted to a higher frequency region as the applied temperature is increased. This shift towards higher frequency regions indicates faster charge transfer possible in the materials.
The plot of real part of impedance vs the imaginary part impedance known as the Cole-Cole plot of Cl-NR and 5 phr composite is exhibited in Fig. 10 at various temperatures. The skewed semi-circular arc is observed for all the Fig. 9 Variation of real and imaginary impedance of Cl-NR and 5 phr NiO filled Cl-NR samples with the presence of a simple tale portion. The diameter of the semi-circular part gives the resistance to charge transfer (R ct ). The radius of curvature of the semicircle is found to decrease with the content of filler, and also with the increase in temperature. Cole-Cole plot for multiwalled carbon nanotube reinforced butyl rubber composites was reported by Dang et al. [48]. The semi-circular plots indicating the presence of capacitive element and the semiconducting behaviour of composite is explained in detail. The charge transfer between the nanotubes and elastomer is observed with the reduction in arc radius as per the report [48]. Cole plots also provide data about characteristic frequency (fc), bulk solution resistance (Rα); actually, the resistance at extremely high frequency. This semi-circular nature of plots clearly explains the dielectric relaxation with a very low relaxation time. Relaxation time obtained from Cole plot can be related to the charge carrier mobility and another dimensionless factor obtained from Cole plot is α which usually lies between 0 and 1 [49]. This shape factor possesses an inverse relation with the widening of the frequency domain in the plot of real impedance and peak spreading in the plot of imaginary impedance. The bulk resistivity of the system is explained by the real axis of the Cole plot and the imaginary axis evaluates the maximum angular frequency of the system. The frequency corresponding to the peak position of the curve is given by the following equation [49].
where Cb and Rb respectively represent the bulk capacitance and bulk resistance. ɩ is the relaxation time for the systems.
The semi-circular curve model of electrical impedance is shown in Fig. 11, where centre of the semicircle with a radius r lies below the X-axis with coordinates (X 0 , Y 0 ). The Cole semicircle is intercepting at R α and R 0 in x axis, The reactance (X) and resistance (R) can be calculated using the following equations [50].
Complex impedance as a function of relaxation time, current density and angular frequency (ω = 2 f) can be shown in the following relation.
The Cole-Cole parameters derived from Cole semi-circular plots are detailed below  As the temperature of the system increases the bulk resistance is found to be decreased and the radius of curvature correspondingly reduced indicating the improved conductivity by the fast-moving charge carriers.
The circuit model proposed for the bulk capacitance and resistance for the composites of Cl-NR/NiO is exhibited in Fig. 12. This can be applied for the fitting of Cl-NR/NiO nanocomposite (with 5 phr filler loading) impedance plot. Where Ra is the resistance within the aggregate of NiO, Rb is the bulk effective resistance and Cb is the constant phase element; also called the bulk capacitance caused by the polarisation at the interface of filler and polymer matrix. The bulk parameters are found to be altered by the effective incorporation of nanofiller, the bulk resistance appreciably reduced and a considerable increase in capacitance is also observed.

Dielectric Modulus
Space charge polarisation and relaxation dynamics of Cl-NR and its composites are evaluated based on their modulus vs frequency plots of them. The complex dielectric modulus is calculated using the following equation [51].
Here, M I = ; real part of modulus and ; imaginary modulus.
The temperature-dependent variation of real and imaginary modulus is well clear from the plots (Fig. 13) and it is found that the modulus is influenced by the temperature to some extent in all the samples analysed. The plots of both real and imaginary parts of modulus are showing a linear frequency-independent region at a lower frequency region indicating the long regime mobility of charge carriers. In the regime of long-range motion of carriers, two possible incidents are present, primarily the poor carrier strength to control their mobility inside the long-range regime and secondly the lack of sufficient restoring force of charge carriers in this long regime [52]. Beyond this characteristic threshold frequency, the real and imaginary modulus is found to increase with frequency. This gradual increase is attributed to the short-range regime of charge movement. The frequency corresponding to the maximum value of imaginary modulus indicates the characteristic relaxation frequency (Table 3). It is found that with temperature the maximum point is shifted towards the higher frequency. The rate of relaxation is increased and thereby the relaxation time decreases. In the case of parent Cl-NR, the constant regime of real and imaginary is found up to log f = 3.5, 2.6 respectively. Beyond this frequency regime, a sharp change is observed in both cases. The plot of imaginary modulus shows a hump in the plot indicating the relaxation process and can confirm the non-Debye type of relaxation possible in these systems. Different studies are focused on the dielectric modulus spectra for analysing the dielectric relaxation, polarisation and Maxwell-Wagner Sillars effect that occurs in the composite systems. The analysis of dielectric loss factor under different temperatures showed a shift of relaxation peak to lower frequencies and broadening of the peak with an increase in ground tire rubber particle loading in LDPE was reported by Genescà et al. [53]. The grain and grain boundary in the electrical conductivity mechanism can be connected to the Bode plot exhibited in Fig. 9 (plot of real and imaginary Z vs log frequency). The usual dielectric behaviour of composite is clear from the high real and imaginary impedance value observed at initial frequencies. This phenomenon can be explained based on the space charge polarisation, that is all the charge carriers possess a very active nature at this low frequency initial portion. Thus, the charge carriers are evenly piled up at the interface of grain boundaries, thereby accumulation of charge carriers occurs at the boundary. Thus, the relaxation phenomenon possible in these dielectric systems can be explained with both these dielectric factors Z and modulus, similar trends were reported in previous literature [54]. The unsystematic  Table 3 can be attributed to the effect of accumulated space charges at the grain boundary interface.

DC Conductivity Modelling Studies
The conductivity of insulating polymers can be enhanced by the incorporated nano metal oxide filler to a great extent, due to the considerable semiconducting nature of the filler particles. The DC conductivity (logarithmic) values obtained for Cl-NR with 0, 1, 3, 5, 7 and 10 phr NiO loaded samples are − 6.74, − 4.75, − 4.74, − 4. 0, − 4.23 and − 4.31 S/ cm respectively. Here the DC conductivity is substantially enhanced upon the effective interpenetration of the metal oxide filler particles into the polymer. The increasing trend is observed up to 5 phr sample, and beyond this threshold filling, there is a slight depreciation in conductivity is due to the formation of lumps by their internal force of attraction exceeding the polymer-filler interactions.
The increase in DC conductivity of the nanocomposites as a function of filler loading is correlated with the different theoretical assumptions based on Scarisbrick, Bueche and McCullough models. The important parameters of the improved conductivity depend on filler-polymer interactions, filler-filler interactions, nature of the fillers and the uniform arrangement of fillers. The function of fillers for the development of conducting networks in the polymer matrix is also analysed by these theoretical models.  Scarisbrick made an assumption on the induced conductivity that the conductivity mainly relies on the arrangement and concentration of filler particles in the polymer matrix [55]. Following fundamental equations are used to analyse this model mathematically.
where the geometric factor C 2 explains the dispersion of filler in the polymer, usually lies between 1 and 3 × 10 −3 . The volume fraction of filler is indicated by Vf, the respective conductivities of filler and polymer are indicated by f , c . The conductivity calculated from Scarisbrick model and the respective comparative graph of experimental DC conductivity and theoretical DC values of composites is exhibited in Fig. 14. It is observed that the experimental plot is relatively lower than the theoretical plots corresponding to different geometric factor values. The specific model is proposed mainly based on the filler particles so that at lower filler loadings it shows a large deviation from the experimental value as the content of the filler is poor at these loadings. As the content of filler increased and at very higher filler contents the proposed model shows rather better concordance. Thus, the large deviation of theoretical values from the experimental DC values nullifies the chance of acceptance of this particular model for the current nanocomposite system based on NiO. Electrical properties of polyvinyl fluoride/ multiwalled carbon nanotube composites reported the ohmic conduction through intermolecular contact of filler particles and the polymer [56] and the present study exhibits the same trend as the Scarisbrick model.
Bueche model explains the different drastic jumps in electrical conductivity and hence it is suitable for binary mixtures. The equation corresponding to the theoretical DC conductivity is given by the following equation [57].
where the theoretical conductivity of composite is indicated by c , the f represents the filler conductivity and Vf is the volume fraction of filler particles. The comparison of experimental and theoretical conductivities is plotted in Fig. 15. The figure explains the large deviation of theoretical values from the respective experimental values indicating the large difference in conductivity of the filler and the bare polymer. This model focused on the addition of conductivity of filler and that of the bare polymer as per Eq. 15. Similar results are also reported by Ram et al. and the results suggest that the theoretical conductivity values are much higher than experimental values [56]. This narrows their application extent to those composites with comparatively similar conductivity counterparts. Thus, this model fails to meet the conductivity variation of Cl-NR/NiO system since NiO is a highly conducting metal oxide filler compared to that of Cl-NR.
McCullough model is used for analysing the transport properties of polymer composites and it can also be used for the conductivity prediction of similar binary systems [58]. The proposed equation for the model is given by the following mathematical expression.
The extent of conductive network formation indicated by the structural factor λ usually lies between 0 and 1. Vf and Vp are expressed by the following equations [43].   Fig. 16. The validation of McCullough model is done with the substitution of different values, it is found that = 0.9 shows the maximum concordant nature with the experimental plot of conductivity. Thus, it can be inferred from the comparison that the McCullough model is a better fit for the conduction mechanism in Cl-NR/NiO system. The conductivity of composites systems based on ethylene-vinyl acetate copolymer at different values was investigated by Sohi et al. [59]. From the obtained results they concluded that the McCullough model lacks in predicting the conductivity of the composite systems as the theoretical conductivity changes is not so pronounced with increasing filler loading.

Conclusion
From the above discussions the major conclusions are: (1) The FT-IR and UV spectra showed the strong interaction of NiO nanoparticles with chlorinated natural rubber through the shift in the position of absorption peak as compared to the control sample. The bandgap energy of the composite was found to decreased with increase in filler loading up to an equilibrium loading level (5phr).
(2) TGA results indicated the improved thermal stability of the composite by the proper insertion of nanofiller in the chlorinated natural rubber. (3) The AC conductivity of composites was increased with an increase in temperature and also with the addition of NiO nanoparticles, which indicated the semiconducting nature of the polymer and the increased conductivity values proved the accelerated conductivity provided by the metal oxide filler particles, activated charge carriers at higher temperatures. (4) The dielectric constant and dielectric loss tangents showed a proportional increase with increasing temperature. Initial higher dielectric values confirmed the electrode polarisation occurred at lower frequency regions and this nature explains the non-Debye type of conduction mechanism possible in the above systems. (5) An appreciable decrease in impedance value was found in the imaginary and real part, explaining the comprehensive depreciation resulting from the hike in conductivity provided by the metal oxide fillers and the applied thermal energy. (6) The radius of curvature of the semicircle in the Cole-Cole plot was found to decrease with the content of filler and also with the increase in temperature. (7) Space charge polarisation and relaxation dynamics of Cl-NR and its composites were evaluated by complex modulus spectra. (8) The DC conductivity predicted by McCullough exhibits somewhat similarity with the experimental conductivity of Cl-NR/NiO systems. (9) The prepared rubber nanocomposites with good optical properties, thermal stability, electrical conductivity and dielectric parameters can be used in flexible electronic appliances, sensors, super-capacitors and actuators.