Effect of ZnO Nanoparticle Content on the Structural and Ionic Transport Parameters of Polyvinyl Alcohol Based Proton-Conducting Polymer Electrolyte Membranes

Proton conducting nanocomposite solid polymer electrolytes (NSPEs) based on polyvinyl alcohol/ammonium nitrate (PVA/NH4NO3) and different contents of zinc oxide nanoparticles (ZnO-NPs) have been prepared using the casting solution method. The XRD analysis revealed that the sample with 2 wt.% ZnO-NPs has a high amorphous content. The ionic conductivity analysis for the prepared membranes has been carried out over a wide range of frequencies at varying temperatures. Impedance analysis shows that sample with 2 wt.% ZnO-NPs has a smaller bulk resistance compared to that of undoped polymer electrolyte. A small amount of ZnO-NPs was found to enhance the proton-conduction significantly; the highest obtainable room-temperature ionic conductivity was 4.71 × 10−4 S/cm. The effect of ZnO-NP content on the transport parameters of the prepared proton-conducting NSPEs was investigated using the Rice–Roth model; the results reveal that the increase in ionic conductivity is due to an increment in the number of proton ions and their mobility.


Introduction
Research on proton-conducting solid polymer electrolytes (SPEs) over the past few decades has aimed to provide high-performance and stable electrochemical devices, such as electrochemical double-layer capacitors, light-emitting electrochemical cells, solid-state batteries, and fuel cells [1][2][3]. The proton transport in SPEs can be designated based on three mechanisms: hopping, diffusion, and transport associated with polymer chain segmental movement [4]. The ion hopping mechanism and ion transport by segmental motions are more favored at higher temperatures [5]. To investigate the proton conduction mechanism of a system, the ionic conductivity is typically characterized in terms of temperature [6].
Nonetheless, the greatest drawback of proton-conducting SPEs is their low ionic conductivity at room temperature, which restricts their practical applications in energy storage devices [7]. In recent years, significant efforts have been dedicated to enhancing ionic conduction in proton-conducting SPEs by different approaches, including polymer blending, copolymerization, the addition of plasticizers, and the incorporation of nanosized inorganic fillers to the system such as carbon nanotubes, reduced graphene oxide, and metal oxide. Among these approaches, the dispersion of a small amount of inorganic nano-sized fillers into the polymer electrolyte matrix has captured escalating interest
The surface morphology of the prepared samples was conducted by scanning electron microscopy (SEM, JEOL JSM-6060) (Tokyo, Japan), operating at 20 kV. The samples were sputtered with thin gold layers prior to imaging.
Electrochemical impedance spectroscopy (EIS) measurements were performed in the frequency range from 100 Hz to 2 MHz and in the temperature range 303 K to 353 K to examine the effect of ZnO-NPs on the ionic conductivity of proton-conducting PVA/NH4NO3 polymer electrolytes. The impedance was measured with a KEYSIGHT E4980A LCR Meter (Santa Rosa, CA, USA) that has been interfaced with a computer. The proton-conducting NSPE samples were mounted on the holder with aluminum blocking electrodes of diameter 1 cm under spring pressure to ensure good contact between NSPE films and electrodes. In this study, the semicircular arc of the Cole-Cole plot of complex
The surface morphology of the prepared samples was conducted by scanning electron microscopy (SEM, JEOL JSM-6060) (Tokyo, Japan), operating at 20 kV. The samples were sputtered with thin gold layers prior to imaging.
Electrochemical impedance spectroscopy (EIS) measurements were performed in the frequency range from 100 Hz to 2 MHz and in the temperature range 303 K to 353 K to examine the effect of ZnO-NPs on the ionic conductivity of proton-conducting PVA/NH 4 NO 3 polymer electrolytes. The impedance was measured with a KEYSIGHT E4980A LCR Meter (Santa Rosa, CA, USA) that has been interfaced with a computer. The proton-conducting NSPE samples were mounted on the holder with aluminum blocking electrodes of diameter 1 cm under spring pressure to ensure good contact between NSPE films and electrodes. In this study, the semicircular arc of the Cole-Cole plot of complex impedance was used to obtain the bulk resistance (R b ), and the electrical conductivity (σ) of the samples was calculated from this equation: Here, l and A are, respectively, the sample thickness and the area of the electrode. A micrometer gauge was used to determine the thickness of films and was found to range from 240 to 315 µm.

XRD Analysis
The XRD patterns for PVA/NH 4 NO 3 /ZnO NSPEs with varying ZnO-NP contents are shown in Figure 2. The XRD patterns for all samples exhibit a broad peak centered at 2θ = 19.25 • , corresponding to the semi-crystalline nature of PVA that arises from the intra-and inter-molecular hydrogen bonding of the O-H groups in the PVA backbone [24]. However, the presence of NH 4  impedance was used to obtain the bulk resistance ( ), and the electrical conductivity ( ) of the samples was calculated from this equation: Here, and are, respectively, the sample thickness and the area of the electrode. A micrometer gauge was used to determine the thickness of films and was found to range from 240 to 315 µm.

XRD Analysis
The XRD patterns for PVA/NH4NO3/ZnO NSPEs with varying ZnO-NP contents are shown in Figure 2. The XRD patterns for all samples exhibit a broad peak centered at 2θ = 19.25°, corresponding to the semi-crystalline nature of PVA that arises from the intraand inter-molecular hydrogen bonding of the O-H groups in the PVA backbone [24]. However, the presence of NH4NO3 crystalline peaks in all samples at 18.07°, 22.61°, 24.46°, 29.10°, 36.24°, and 40.38° indicates the presence of some undissociated NH4NO3 salt in the NSPE samples [25]. The minimum intensity of the characteristic peaks upon loading 2 wt.% ZnO-NP reveals the lowest relative crystallinity of the NSPE-2 sample. Beyond this concentration, the intensity of the NH4NO3 peaks increases as the ZnO-NP concentration increases, which causes a decrement in the conductivity due to the recombination of the dissociated ions to form NH4NO3 salt. It has been well reported that even a small change in the crystallinity of polymer samples has a profound effect on the conductivity [26][27][28]. Figure 3 depicts the scanning electron microscope (SEM) micrographs of PVA/NH4NO3/ZnO NSPEs with different ZnO-NP contents. It can be seen that the morphology of undoped and doped proton-conducting polymer electrolytes consists of solid structures that have protruded out of the membrane surface, revealing that the NH4NO3 salt has recrystallized out of the NSPE surface, which is in good agreement with the XRD results. It is also clear that the density of the solid structures reduced after the addition of 2 wt.% of ZnO-NPs. A further increase in ZnO-NP content resulted in a further increase The minimum intensity of the characteristic peaks upon loading 2 wt.% ZnO-NP reveals the lowest relative crystallinity of the NSPE-2 sample. Beyond this concentration, the intensity of the NH 4 NO 3 peaks increases as the ZnO-NP concentration increases, which causes a decrement in the conductivity due to the recombination of the dissociated ions to form NH 4 NO 3 salt. It has been well reported that even a small change in the crystallinity of polymer samples has a profound effect on the conductivity [26][27][28]. Figure 3 depicts the scanning electron microscope (SEM) micrographs of PVA/NH 4 NO 3 / ZnO NSPEs with different ZnO-NP contents. It can be seen that the morphology of undoped and doped proton-conducting polymer electrolytes consists of solid structures that have protruded out of the membrane surface, revealing that the NH 4 NO 3 salt has recrystallized out of the NSPE surface, which is in good agreement with the XRD results. It is also clear that the density of the solid structures reduced after the addition of 2 wt.% of ZnO-NPs. A further increase in ZnO-NP content resulted in a further increase in the size and density of these solid structures in the polymer electrolytes. This observation is in agreement with the aforementioned XRD analysis. The profile plots show a part of a depressed semicircle for all samples. The semicircle arc symbolizes the parallel combination of bulk resistance (due to mobile ions inside the polymer matrix) and bulk capacitance (due to immobile polymer chains) [29]. As the ZnO content increases, the semicircle in the plots was observed to lessen up to 2 wt.%; beyond this concentration, the trend is reversed.

Impedance Analysis
Membranes 2021, 11, x FOR PEER REVIEW 5 of 14 in the size and density of these solid structures in the polymer electrolytes. This observation is in agreement with the aforementioned XRD analysis.  Figure 4 represents the Cole-Cole plots of complex impedance for samples with 0, 1, 2, 3, 4, and 5 wt.% ZnO-NPs loaded PVA/NH4NO3 polymer electrolytes at room temperature. The profile plots show a part of a depressed semicircle for all samples. The semicircle arc symbolizes the parallel combination of bulk resistance (due to mobile ions inside the polymer matrix) and bulk capacitance (due to immobile polymer chains) [29]. As the The R b for NSPE samples has been found from the intercept of the semicircle arc at low-frequency on the real Z axis. As shown in the inset of Figure 4, the R b has been found to be 6.92 × 10 6 , 7.36 × 10 5 , 1.79 × 10 5 , 7.51 × 10 5 , 5.64 × 10 6 , and 1.15 × 10 7 Ω for 0, 1, 2, 3, 4, and 5 wt.% content of ZnO-NPs, respectively.

Impedance Analysis
The conductivity of proton-conducting PVA/NH 4 NO 3 /ZnO NSPE membranes was observed to increase with temperature. At high temperatures, the dissociation of ammonium nitrate and the thermal movement of PVA molecular chain segments would be improved, which caused an increase in the ionic conductivity.
The complex-plane impedance plots of the NSPE-2 sample at different temperatures are presented in Figure 5. It is obvious that, as the temperature increased, the depressed semicircular arc in the plots was observed to lessen and finally disappear, leaving only a low-frequency spike. This suggests the existence of only the resistive component [29,30], which reveals the absence of capacitive nature; therefore, only the diffusion processes take place at high temperatures [31]. ZnO content increases, the semicircle in the plots was observed to lessen up to 2 wt.%; beyond this concentration, the trend is reversed. The for NSPE samples has been found from the intercept of the semicircle arc at low-frequency on the real ′ axis. As shown in the inset of Figure 4, the has been found to be 6.92 × 10 6 , 7.36 × 10 5 , 1.79 × 10 5 , 7.51 × 10 5 , 5.64 × 10 6 , and 1.15 × 10 7 Ω for 0, 1, 2, 3, 4, and 5 wt.% content of ZnO-NPs, respectively.
The conductivity of proton-conducting PVA/NH4NO3/ZnO NSPE membranes was observed to increase with temperature. At high temperatures, the dissociation of ammonium nitrate and the thermal movement of PVA molecular chain segments would be improved, which caused an increase in the ionic conductivity.
The complex-plane impedance plots of the NSPE-2 sample at different temperatures are presented in Figure 5. It is obvious that, as the temperature increased, the depressed semicircular arc in the plots was observed to lessen and finally disappear, leaving only a

Conductivity Analysis
The variation in room-temperature direct current (DC) conductivity (σ DC ) versus ZnO-NP content is presented in Figure 6. The dependence of σ DC on ZnO content provides information on the particular interaction between ions of NH 4 NO 3 salt and the functional group of the PVA matrix. Kadir et al. [32] reported that the chitosan-PVA-NH 4 NO 3 proton-conducting system gives the optimum value of ionic conductivity 2.07 × 10 −5 S/cm at room temperature upon incorporating 40 wt.% salt, which is comparable with the undoped SPE sample in the present work. The peak observed in Figure 6 depicts the roomtemperature highest DC conductivity optimized at 4.71 × 10 −4 S/cm with the addition of 2 wt.% of ZnO-NPs. Beyond that, the ionic conductivity decreases quickly. Tripathi and Kumar [33] attributed the decrease in conductivity with increasing ZnO concentration beyond 3 wt.% in plasticized polymer gel electrolytes based on poly(vinylidene fluoride)co-hexafluoropropylene (PVDF-HFP) to the small value of ZnO-NPs dielectric constant compared to the polymer gel electrolyte system. low-frequency spike. This suggests the existence of only the resistive component [29,30], which reveals the absence of capacitive nature; therefore, only the diffusion processes take place at high temperatures [31].

Conductivity Analysis
The variation in room-temperature direct current (DC) conductivity ( ) versus ZnO-NP content is presented in Figure 6. The dependence of on ZnO content provides information on the particular interaction between ions of NH4NO3 salt and the functional group of the PVA matrix. Kadir et al. [32] reported that the chitosan-PVA-NH4NO3 proton-conducting system gives the optimum value of ionic conductivity 2.07 × 10 −5 S/cm at room temperature upon incorporating 40 wt.% salt, which is comparable with the undoped SPE sample in the present work. The peak observed in Figure 6 depicts the roomtemperature highest DC conductivity optimized at 4.71 × 10 −4 S/cm with the addition of 2 wt.% of ZnO-NPs. Beyond that, the ionic conductivity decreases quickly. Tripathi and Kumar [33] attributed the decrease in conductivity with increasing ZnO concentration beyond 3 wt.% in plasticized polymer gel electrolytes based on poly(vinylidene fluoride)co-hexafluoropropylene (PVDF-HFP) to the small value of ZnO-NPs' dielectric constant compared to the polymer gel electrolyte system. The average value of for all prepared samples is tabulated in Table 1. The increases in with the addition of ZnO-NPs could be related to the increase in both the number and mobility of free carriers in the matrix by increasing the degree of salt dissociation of ion aggregates, and increasing the amorphous phase content, respectively [34,35]. The average value of σ DC for all prepared samples is tabulated in Table 1. The increases in σ DC with the addition of ZnO-NPs could be related to the increase in both the number and mobility of free carriers in the matrix by increasing the degree of salt dissociation of ion aggregates, and increasing the amorphous phase content, respectively [34,35]. The temperature-dependence of ionic conductivity has been employed to analyze the possible ion-conduction mechanism in the present proton-conducting NSPEs. Figure 7 shows the plot of log σ DC versus 1000/T for different ZnO-NP contents in PVA/NH 4 NO 3 polymer electrolyte membranes. The linear variation of these plots suggests that the thermally activated process exhibits the Arrhenius-type behavior [36]. However, the observed linear relations for all doped PVA/NH 4 NO 3 polymer electrolyte samples mean that there is no phase transition marked in the NSPEs by adding ZnO-NPs. As per this model, the temperature-dependent σ DC can be expressed by activation energy (E A ), which is obtained in terms of the Arrhenius equation [37]: where σ o and k B represent the pre-exponential factor and Boltzmann constant, respectively. The value of E A was calculated using the grade of the Arrhenius plot shown in Figure 7. The increase in the of NSPEs with temperature can be understood as the hopping of proton ions between PVA coordinating sites; hopping being helped by both polymer chain segmental motions and local structural relaxations [38]. As the temperature increases, the amorphous domains gradually increase, and the polymer chains earn faster internal modes producing segmental motion due to bond rotations. As a result, ion hopping due to inter-and intra-chain movements is favored, causing the spectacular to enhance the conductivity of the matrix [39,40]. The calculated values of and in accordance with ZnO content are presented in Table 1, which also shows that the is in- The increase in the σ DC of NSPEs with temperature can be understood as the hopping of proton ions between PVA coordinating sites; hopping being helped by both polymer chain segmental motions and local structural relaxations [38]. As the temperature increases, the amorphous domains gradually increase, and the polymer chains earn faster internal Membranes 2021, 11, 163 9 of 13 modes producing segmental motion due to bond rotations. As a result, ion hopping due to inter-and intra-chain movements is favored, causing the spectacular to enhance the conductivity of the matrix [39,40]. The calculated values of σ DC and E A in accordance with ZnO content are presented in Table 1, which also shows that the E A is inversely proportional to the σ DC in the manner that the highest conductivity sample (NSPE-2) shows the minimum value of hopping activation energy. Here, the E A has been assigned as the energy acquired by the H + ion to free itself from its localized-state. This reveals that incorporating a small amount (2 wt.%) of ZnO-NPs into the PVA/NH 4 NO 3 polymer electrolyte causes a reduction in the potential energy barriers for the proton transport, leading to a decrease in the activation energy [41]. This result is expected and comparable with much previous work for different proton-conducting polymer electrolyte systems [29,34,42]. Hema et al. [43] also observed that the temperature-dependent conductivity for the proton-conducting polymer electrolyte based on PVA-NH 4 Cl, PVA-NH 4 Br, and PVA-NH 4 I followed the Arrhenius-type relationship.

Dielectric Study
The dielectric study of the present proton-conducting NSPEs was carried out to understand the conductivity behavior of the systems and is explained in terms of the real (M ) and imaginary (M ) parts of electric modulus, as they are free from the contribution of the interfacial electrode/electrolyte polarization effect at low frequencies. The dielectric study gives information on relaxing dipoles in the samples [44]. The obtained complex permittivity (ε * ) data were analyzed using complex modulus (M * ), which is an inverse of the ε * and is linked to the impedance data as follows: Here, ω is the angular frequency (ω = 2π f , f being frequency), C o = ε o A/d, where ε o is the free space permittivity, A is the cross-section of the electrode, and d is the film thickness, and Z * is the complex impedance.
The variations of real (M ) and imaginary (M ) parts of electric modulus for the PVA/NH 4 NO 3 doped 2 wt.% ZnO at various temperatures are introduced in Figure 8. The plot of M and M shows low value at lower frequencies, which is caused by the huge value of interfacial capacitance correlated with the electrode-electrolyte boundary [34]. However, no definitive peaks can be observed for the M plot, and the M spectrum shows an asymmetry relaxation peak accompanied by the dispersion of M in the frequency range employed in this study. The broadness and asymmetry shape of the M peak disclose the distribution of relaxation time and non-Debye relaxation process [14,45].
A shift in the M relaxation peak towards the higher-frequency side with a temperature rise indicates the reduction in the relaxation time, which directly supports the ionic conductivity enhancement as a consequence of an increase in the mobility of free ions [17]. According to Khiar and Arof [46], as temperature increases, the degree of salt dissociation and re-dissociation of ion aggregates causes an increase in the number of free ions.
The combined plots of Z and M against frequency are usually used to identify whether the short-range or long-range motion of free carriers is dominant in the relaxation process. The mismatch of frequency peaks between Z and M reflects that the shortrange movement of free carriers is the predominant process and departs from the ideal Debye-type model, whereas the coincidence of the frequency peaks at a similar frequency implies that the long-range movement of free carriers is dominant [47]. The frequency response of normalized Z /Z max and M /M max for the sample with 2 wt.% of ZnO-NPs at temperatures 30 and 60 • C was represented in Figure 9. From this figure, it is noticed that the Z /Z max and M /M max peaks do not concur, indicating the short-range movement of free carriers and non-Debye relaxation processes in the present proton-conducting NSPE sample. The mismatch between the Z /Z max and M /M max peaks become larger with increasing temperature, which suggests the increases in the portion of the short-range movement of free carriers with increasing temperature. This result is in accordance with previous works, which suggest that the conductivity increases with increasing temperature.
The variations of real ( ) and imaginary ( ) parts of electric modulus for the PVA/NH4NO3 doped 2 wt.% ZnO at various temperatures are introduced in Figure 8. The plot of and shows low value at lower frequencies, which is caused by the huge value of interfacial capacitance correlated with the electrode-electrolyte boundary [34]. However, no definitive peaks can be observed for the ′ plot, and the ′′ spectrum shows an asymmetry relaxation peak accompanied by the dispersion of ′ in the frequency range employed in this study. The broadness and asymmetry shape of the ′′ peak disclose the distribution of relaxation time and non-Debye relaxation process [14,45]. A shift in the ′′ relaxation peak towards the higher-frequency side with a temperature rise indicates the reduction in the relaxation time, which directly supports the ionic conductivity enhancement as a consequence of an increase in the mobility of free ions [17]. According to Khiar and Arof [46], as temperature increases, the degree of salt dissociation and re-dissociation of ion aggregates causes an increase in the number of free ions. The combined plots of ′′ and ′′ against frequency are usually used to identify whether the short-range or long-range motion of free carriers is dominant in the relaxation process. The mismatch of frequency peaks between ′′ and ′′ reflects that the shortrange movement of free carriers is the predominant process and departs from the ideal Debye-type model, whereas the coincidence of the frequency peaks at a similar frequency implies that the long-range movement of free carriers is dominant [47]. The frequency response of normalized / ′′ and / ′′ for the sample with 2 wt.% of ZnO-NPs at temperatures 30 and 60 °C was represented in Figure 9. From this figure, it is noticed that the / ′′ and / ′′ peaks do not concur, indicating the short-range movement of free carriers and non-Debye relaxation processes in the present proton-conducting NSPE sample. The mismatch between the / ′′ and / ′′ peaks become larger with increasing temperature, which suggests the increases in the portion of the short-range movement of free carriers with increasing temperature. This result is in accordance with previous works, which suggest that the conductivity increases with increasing temperature.

Ion Transport Parameters
Ion transport parameters such as charge carrier density ( ), its mobility ( ), and the diffusion coefficient ( ) of the present proton-conducting PVA/NH4NO3/ZnO NSPE membranes are investigated in detail using the Rice-Roth model [48]. This model postulated that the ionic carrier of mass in the localized states could be thermally excited to a free-ion-like state after receiving energy equal to the activation energy of conduction ( ); wherein the ion is propagated through the electrolyte with a velocity ( ) [49], given by: = 2 / . The mean free path of ion transport or the distance traveled by the ion between two complexation sites (ℓ) is given as: ℓ = , where is the time of ions' travel from one complex site to another. In the present work, ℓ is the hopping distance between two repeating units of hydroxyl groups in PVA, which is taken to be around 2.15 Å [50,51].

Ion Transport Parameters
Ion transport parameters such as charge carrier density (n), its mobility (µ), and the diffusion coefficient (D) of the present proton-conducting PVA/NH 4 NO 3 /ZnO NSPE membranes are investigated in detail using the Rice-Roth model [48]. This model postulated that the ionic carrier of mass m in the localized states could be thermally excited to a free-ion-like state after receiving energy equal to the activation energy of conduction (E A ); wherein the ion is propagated through the electrolyte with a velocity (v) [49], given by: The mean free path of ion transport or the distance traveled by the ion between two complexation sites ( ) is given as: = vτ, where τ is the time of ions travel from one complex site to another. In the present work, is the hopping distance between two repeating units of hydroxyl groups in PVA, which is taken to be around 2.15 Å [50,51].
According to the Rice-Roth model, the ionic conductivity of free mobile ion is expressed as: Here, Z is the valency of the conducting ions, e is the electron charge, and k B is the Boltzmann constant. Equation (4) was used to evaluate the number density of mobile ions (n). From the estimated value of n, the ionic mobility (µ = σ/ne) and diffusion coefficient (D = k B Tσ/ne 2 ) of the samples can also be calculated. Table 1 shows the value of n, µ, and D for the PVA/NH 4 NO 3 /ZnO NSPEs samples with different ZnO contents. It is seen from the table that the maximum conducting sample has a maximum value of n and µ, which confirms that the conductivity in the present NSPEs is actually controlled by both the number and mobility of H + ions in the samples. These studies indicate that the conductivity of the PVA/NH 4 NO 3 polymer electrolyte can be enhanced moderately by adding a small percentage of ZnO-NPs, owing to the increase in both the mobility and number density of mobile proton ions.

Conclusions
Proton-conducting PVA/NH 4 NO 3 /ZnO NSPE membranes with different contents of ZnO-NPs were prepared using the cast technique. The small percentage of ZnO-NPs was found to influence the proton-conduction of the system, and the highest obtained value of conductivity is 4.71 × 10 −4 S/cm at room temperature. The temperature-dependent ionic conductivity results exhibited Arrhenius behavior, and the activation energy values were inversely proportional to the DC conductivity. Dielectric studies suggest that the NSPE samples in this study exhibit non-Debye behavior, and the relaxation process is caused by the short-range movement of free carriers. The application of the Rice-Roth model deduced that the increase in conductivity arose from the increase in the mobility and number density of mobile proton ions in the system.

Data Availability Statement:
The authors confirm that the data supporting the findings of this study are available within the article.