Electrocatalytic and Enhanced Photocatalytic Applications of Sodium Niobate Nanoparticles Developed by Citrate Precursor Route

Development of cost effective and efficient electrocatalysts is crucial to generate H2 as an alternative source of energy. However, expensive noble metal based electrocatalysts show best electrocatalytic performances which acts as main bottle-neck for commercial application. Therefore, non-precious electrocatalysts have become important for hydrogen and oxygen evolution reactions. Herein, we report the synthesis of high surface area (35 m2/g) sodium niobate nanoparticles by citrate precursor method. These nanoparticles were characterized by different techniques like X-ray diffraction, transmission electron microscopy and X-ray photoelectron spectroscopy. Electrocatalytic properties of cost-effective sodium niobate nanoparticles were investigated for HER and OER in 0.5 M KOH electrolyte using Ag/AgCl as reference electrode. The sodium niobate electrode showed significant current density for both OER (≈2.7 mA/cm2) and HER (≈0.7 mA/cm2) with onset potential of 0.9 V for OER and 0.6 V for HER. As-prepared sodium niobate nanoparticles show enhanced photocatalytic property (86% removal) towards the degradation of rose Bengal dye. Dielectric behaviour at different sintering temperatures was explained by Koop’s theory and Maxwell-Wagner mechanism. The dielectric constants of 41 and 38.5 and the dielectric losses of 0.04 and 0.025 were observed for the samples sintered at 500 °C and 700 °C, respectively at 500 kHz. Conductivity of the samples was understood by using power law fit.


Results and Discussion
structural characteristics. The phase composition, crystallinity and purity of the samples was analysed by X-Ray diffraction technique. The XRD result of the precursor obtained at 300 °C is given in Fig. S1. All the observed diffraction peaks of the precursor correspond to the standard JCPDS card No. 731788 having chemical composition of NaNb 13 O 33 . The XRD results of NaNbO 3 nanoparticles obtained at 500 °C is shown in Fig. 1a. From XRD profile, seven diffraction peaks appear at 22.88, 32.42, 39.79, 46.57, 52.38, 57.79 and 68.01 degrees, were indexed to (101), (121), (220), (202), (141), (123) and (242) lattice planes, respectively. It was revealed that the diffraction peaks observed in XRD correspond to the NaNbO 3 having orthorhombic phase, with lattice parameters (a = 5.569 Å, b = 7.790 Å, c = 5.518 Å) and space group P2 1 ma (JCPDS No. 742454). The XRD pattern shows that the synthesized NaNbO 3 nanoparticles are highly crystalline and no phase other than orthorhombic NaNbO 3 have been formed. Bulk NaNbO 3 was also synthesized to compare the electrocatalytic properties. The XRD of bulk NaNbO 3 is shown in Fig. 1b, which matches with NaNbO 3 (JCPDS No. 895173) possessing orthorhombic phase structure. The intensity of reflections appeared for bulk NaNbO 3 is found to be high as compared to NaNbO 3 nanoparticles synthesized by polymeric citrate precursor route, which is due to high crystallinity and large particle size of the sample. electron microscopic studies. The electron microscopic study was carried out by using TEM technique. A uniform dispersion of synthesized samples in water for TEM analysis was prepared by ultrasonication for 20 min and sample preparation was done by drop casting of a dispersed sample on copper grid. Figure 2a represents the TEM micrograph of as-synthesized NaNbO 3 . Figure 2b represents the size distribution histogram of the as-synthesized nanoparticles. TEM micrograph shows that nanoparticles having size in the range of 5-30 nm were synthesized and average size of particles was found to be 15 nm. Therefore, the TEM result shows that synthesis of NaNbO 3 could be effectively controlled to nanoscale level using polymeric precursor method. TEM image also shows that the non-uniformly synthesized nanoparticles have different geometries but, in TEM image orthorhombic shaped nanoparticles with a ≠ b ≠ c is also apparently visualized in the inset of Fig. 2a. The lattice fringes visible in HRTEM image could be ascribed to NaNbO 3 and the d spacing was found to be 0.178 nm which corresponds to the (141) crystal plane of NaNbO 3 nanoparticles as shown in Fig. 2c. TEM micrograph of bulk NaNbO 3 particles as shown in Fig. 2d suggests large particle size as compared to the nanoparticles synthesized by PCP route. X-ray photoelectron spectroscopic (XPS) analysis. XPS was employed to evaluate the chemical state and chemical composition of as-prepared NaNbO 3 photocatalyst. Figure 3a represents the wide scan XPS spectrum of NaNbO 3 nanoparticles. From XPS, it was observed that synthesized nanoparticles consist of Nb, Na, O and C. Full range XPS spectra show peaks for Na-1s, O-1s, Nb-3d and C-1s. Figure 3b displays the high resolution specific XPS spectra of Nb 3d and shows two signals positioned at 212.6 and 215.5 eV. These peaks correspond to 3d doublet of Nb at 212.6 for Nb 3d 5/2 and at 215.5 eV for Nb 3d 3/2 . From peaks it is clear that there is no satellite peak present with Nb which is present in +5 chemical state in the synthesized material. The high resolution XPS spectra for Na as shown in Fig. 3c, shows the peak at 1077.4 eV which demonstrates that Na 1s exists as single peak which indicates that Na ions have +1 chemical state in the prepared nanoparticles. Similarly, the high-resolution peak for O-1s is shown in Fig. 3d. Only one single peak is observed in high resolved XPS spectra of oxygen showing that O is present in single atmosphere in as-prepared compound. The peak observed at 535.8 eV showing that O is attached to Nb and Na ions in NaNbO 3 .
BET surface area analysis. Surface area is an important factor which influences the electrocatalytic and photocatalytic activities of the catalysts. Therefore, before discussing the electrocatalytic and photocatalytic properties of the synthesized sample, analysis of surface area is of primary importance. In general, larger the surface area of the catalyst greater is its catalytic activity. This is because, materials with high surface area offer more surface-active sites which leads to increase in adsorption and reaction sites for catalytic process, leading to an improvement in photocatalytic and electrocatalytic processes. Large surface area besides increasing the active www.nature.com/scientificreports www.nature.com/scientificreports/ sites for adsorption but reduces the length for ionic diffusions thus, provides the kinetically suitable structures for electrocatalysis. To study the surface area and pore size of the sample, N 2 adsorption-desorption measurements were done using multipoint BET method. Figure 4a represents the N 2 adsorption-desorption curves of NaNbO 3 nanoparticles, which follows the type III isotherm indicating that additional adsorption is taking place due to strong interaction between the adsorbent and the adsorbed layer as compared to interaction between adsorbate and the adsorbent surface i.e. multilayer adsorption is taking place in the adsorbed molecules clustered at more favourable sites 51 . Hysteresis of such type is generally formed by the solids consisting of aggregated or agglomerated particles which lead to the formation of slit shaped pores having nonuniform size or shape. The specific surface area for NaNbO 3 nanoparticles was found to be 35 m 2 /g, which is higher than the data present in the literature [52][53][54] . High surface area is attributed to the reduction in size to nanoscale range. Dubinin-Astakhov (DA) pore size distribution of NaNbO 3 was deduced from the plot using both adsorption and desorption data points as shown in Fig. 4b. The distribution of pore size was found to be 7.5-22.5 Å and the onset of the plot was found to be at 11.5 Å, which confirms the mesoporous characteristic of the sample. From the adsorption points of the sample, the average pore size distribution was determined by Barrett-Joyner-Halenda (BJH) plot as shown in Fig. 4c. Derived from the calculations using BJH, the sample shows wide average pore size distribution centred at ≈58 Å. The results obtained show that mesoporosity is enhanced using citrate precursor route, particularly due to the pre-occupied space by precursor. Although pre-occupied space by precursor is not directly involved in the pore formation therefore it affects the porosity of the nanoparticles as reported by Chen et al. 55 which is attributed to the increased surface area of the material. oeR catalytic performance of sodium niobate. Our primary interest in alkali niobates is to study their electrocatalytic behaviour for OER at room temperature. The nanosized NaNbO 3 should exhibit certain degree of catalytic property due to its high surface area and porous nature as confirmed from BET studies, which could allow the ionic species to diffuse easily and hence promotes effective consumption of active sites. Figure 5a shows the cyclic voltammetry plot of NaNbO 3 nanoparticles. Nanosized NaNbO 3 start generating current at 1.09 V vs Ag/AgCl at a scan rate of 100 mV/s and above this potential it acts as electrocatalyst for OER. The amount of oxygen evolved during the electrocatalytic process is directly proportional to the intensity of the anodic peak current. www.nature.com/scientificreports www.nature.com/scientificreports/ Linear sweep measurements for OER were carried out at a scan rate of 100 mV/s by sweeping the potential across negative to positive value ranging from −0.2 to 1.8 V. Figure 5b represents a typical polarization plot of the sodium niobate electrocatalyst, which exhibits significant OER activity with onset potential of 0.9 V with respect to Ag/AgCl electrode and above this potential there occurs an abrupt rise in the anodic current as a result of O 2 evolution. The obtained onset potential is comparable to the art-of-state electrocatalysts which include RuO 2 (≈1.45 V), IrO 2 (≈1.50 V) and double perovskite-based catalyst like Ba 0.5 Sr 0.5 Co 0.8 Fe 0.2 O 3 (BSCF ≈1.50 V), Pr 0.5 Ba 0.5 CoO 3 (≈1.50 V) 26,56-63 . Nanosized NaNbO 3 show significant electrocatalytic activity (2.7 mA/cm 2 in alkaline medium) compared to IrO 2 and RuO 2 which shows current density of (~4 µA/cm 2 in acidic and ~2 µA/cm 2 in basic electrolyte solution) and (~10 μA/cm 2 in acidic and ~3 μA/cm 2 in alkaline electrolyte solution) respectively 58 . To compare the activity of nanosized high surface area NaNbO 3 , the electrocatalytic activity of bulk NaNbO 3 with large particle size was also studied. Figure S2a represents the LSV plot of bulk NaNbO 3 . LSV plot shows very small current was generated (0.000004 mA/cm 2 ) from bulk NaNbO 3 surface, which confirms that NaNbO 3 particles having small size and high surface area show better electrocatalytic activity compared to bulk NaNbO 3 . More insight into the OER activity was gained by studying the kinetic performance of sodium niobate for OER activity by employing the tafel plot, which is obtained from the polarization curve using the tafel equation where η, b, J and J o corresponds to the over potential, tafel slope, current density and exchange current density of the electrocatalytic reaction respectively. The tafel plot of sodium niobate nanoparticles and bulk NaNbO 3 is shown in Figs 5c and S2b. The tafel slope value for nanosized and bulk NaNbO 3 was equal to 370 and 409.6 mV/ decade respectively. The resulting tafel slope value of NaNbO 3 synthesized by PCP route is comparable to Pt based electrocatalyst 61 , but larger than the standard electrocatalysts like RuO 2 and IrO 2 62,63 , whereas for bulk NaNbO 3 tafel slope is too high showing slow reaction rate for OER. For long term functioning of the electrode, the stability of the material is much more crucial. Herein, the stability of NaNbO 3 electrode material was checked by Cyclic voltammetry at 100 mV/s upto 50 cycles at a potential window ranging from 0 to 1.65 V. Figure 5d shows that the catalyst displays little loss of activity, confirming the stability of the electrode material.
The catalytic activity of NaNbO 3 nanoparticles towards HER was also investigated by casting 0.24 mg/cm 2 of sample on the electrode surface at room temperature and saturated alkaline (0.5 M KOH) electrolyte solution was used to carry out analysis. Figure 6a shows the CV measurements of NaNbO 3 nanoparticles for HER activity, carried out in a potential window ranging from 0 to −1.4 V versus Ag/AgCl at a scan rate of 100 mV/s. The peak which is observed at ~−0.34 V could be due to the reduction of oxygen during the cathodic sweep of CV. www.nature.com/scientificreports www.nature.com/scientificreports/ Thereafter, hydrogen evolution reaction starts by the electrolysis of water. From Fig. 6a, it can be visualized that at 100 mV/s scan rate nanosized sodium niobate shows HER activity according to the reaction below. Figure 6b shows the linear sweep polarization curve of the sodium niobate nanoparticles for HER. The polarization curve shows that, sodium niobate nanoparticles show onset potential equal to ~−1.0 V with respect to Ag/ AgCl in alkaline medium, which is comparable to the already reported electrocatalysts 64 , but the current density obtained is low and therefore, less hydrogen is produced during the HER process. Similarly, Fig. S3a represents the LSV plot of bulk NaNbO 3 sample. A very small current was observed as compared to nanosized NaNbO 3 , which demonstrates that small amount of hydrogen was evolved from the bulk catalyst surface. Table 1 shows the comparison of electrocatalytic OER and HER activity of NaNbO 3 nanoparticles and bulk sample. We have used Nernst equation to convert the potential of Ag/AgCl electrode to the potential of RHE using the i.e.
A small difference in the current density (i.e. 0.1 mA/cm 2 ) was observed from CV and LSV. We assume that this small difference could be due to the experimental handling error in the preparation of electrode. Note that we have prepared fresh electrodes for each experiment. Reaction kinetics and mechanism for electrocatalytic HER was studied using tafel equation. Figures 6c and S3b shows the tafel plot of electrocatalytic HER on nano and bulk NaNbO 3 catalyst surface respectively and the electrocatalytic kinetics for HER was determined by curve fitting of tafel plot. Figures 6c and S3b give the tafel slope values of 113 and 347 mV/decade for electrocatalytic HER over nano and bulk electrocatalysts respectively, which predicts the possible additional water dissociation step in HER on NaNbO 3 nanoparticles. The tafel slope obtained for HER is comparable to the standard electrocatalysts in alkaline medium and is less as compared to many non-noble metal electrocatalysts 65 . photocatalytic properties. The photocatalytic activity of the synthesized NaNbO 3 nanoparticles was estimated by carrying out the degradation of RB organic dye by employing solar radiations as light source and efficiency of photocatalytic activity of as-prepared sample was also examined. To carry out the actual degradation experiment, several control experiments were done to determine the effect of sunlight and surface adsorption of catalysts without light source. Generally, catalytic surfaces are responsible for the photo-oxidation reaction; therefore, the adsorption phenomenon of catalyst/organic dye/pollutant is an important factor for photocatalysis. The ability of NaNbO 3 to adsorb RB dye on its surface was evaluated in absence of sunlight. The experiment www.nature.com/scientificreports www.nature.com/scientificreports/ revealed that no evident adsorption is taking place over the NaNbO 3 photocatalyst surface. The adsorption profile of the synthesized nanoparticles in absence of light irradiation is shown in Fig. S4. From the adsorption study of the nanoparticles, it was observed that after 100 min very little amount of dye has been adsorbed by NaNbO 3 nanoparticles.
Efficiency of synthesized photocatalyst was evaluated by decolourization of 1 × 10 −5 M concentration RB dye solution for 80 min. The decolourization of RB dye was also tested in absence of the photocatalyst. From the experiments, it was determined that the RB dye does not undergo auto degradation in 80 min at pH 7.2. After 30 min in dark, the dye solution was exposed to bright sunlight and degradation was evaluated by monitoring the change in intensity of λ max at 545 nm after every 10 min by using UV-vis spectrophotometer. Figure 7a shows UV-vis spectra of degradation of RB dye by NaNbO 3 photocatalyst under sunlight. The relative intensity of λ max of RB dye molecule continuously decreases with respect to time, which confirms that NaNbO 3 photocatalysts is carrying out the degradation of RB dye. From the photocatalytic dye degradation study, it was confirmed that the synthesized sample is an active photocatalyst. Figure 7b shows the relative change in concentration (C/C 0 ) of RB dye in presence of NaNbO 3 catalyst with time. The calculations for percentage removal of dye by NaNbO 3 photocatalysts were carried out by using equation as discussed in experimental section. Figure 7c demonstrates the RB dye removal percentage from aqueous solution and it was found that NaNbO 3 eliminates ≈86% of dye in 80 mins when exposed to sunlight which is much higher than NaNbO 3 /ZnO heterojunction photocatalyst (75%), pristine NaNbO 3 (55%) and TiO 2 (60%) [65][66][67] .
The well-known factors which influence the ABO 3 photocatalytic performance are structure (anionic and cationic sites), surface area, band gap, defect density and redox ability of ions present at B-site (lattice oxygen activity). In comparison to the bulk NaNbO 3 , an important reason for enhanced photocatalytic activity of NaNbO 3 under solar irradiation in this work is due to the large surface area of NaNbO 3 nanoparticles. The increased surface area results in production of more reaction active sites which enhances the reaction rate of the nanocatalyst.
In perovskite metal oxides, A cation sites are usually considered as catalytically inert as compared to the B-site cations and consequently, A-site cations have marginally less effect on photocatalysis as compared to the cations at site B, whose valence band and conduction band comprises of O 2p and B 3d orbitals respectively 68,69 .
It can be concluded that NaNbO 3 based perovskite catalyst shows considerable photocatalytic activity in degradation of RB dye when irradiated with solar radiations, due the formation of NbO 6 octahedral chains which could favour the possible delocalization of charge carriers during the catalytic reaction. To determine the active www.nature.com/scientificreports www.nature.com/scientificreports/ species and possible photocatalytic degradation mechanism responsible for removal of organic dye, different scavengers such as Benzoquinone (BQ), isopropanol (IP), silver nitrate (AgNO 3 ) and ammonium oxalate (AO) were used. BQ, IP, AgNO 3 and AO acts as scavengers for O 2 •− , OH • , e − and h + respectively. Scavengers were employed to quench different reactions carried out by different oxidising species during photocatalysis. As a result of quenching, the more photocatalytic activity is reduced by any scavenger, the more actively is the oxidising species taking part in degradation reaction. Figure S5 represents the reduction in the photocatalytic activity of the catalyst by using different scavengers. From the quenching results, it was observed that addition of BQ, IP and AgNO 3 reduces the catalytic activity of photocatalyst to large extent. While as, addition of AO has least effect on the catalytic properties of the catalyst. The results obtained confirms the predominant role of OH • in photocatalytic reaction along with photogenerated electrons and superoxide radical anion. The possible mechanism for photocatalytic degradation over the surface of NaNbO 3 is described by following equations.   The conduction band electrons of NaNbO 3 reduce the molecular oxygen to the hydroperoxyl radicals according to reaction below: Under solar irradiation, the valence band electrons are excited and are transferred to its conduction band thereby, results in the creation of photoactive electron and hole pairs. The photogenerated electrons can be further used by the molecular oxygen to excite molecular oxygen for formation of activated species like superoxide 70 thus, results in the reduced electron and hole recombination by prolonging their lifetime. The generated superoxide radical anion helps in the formation of hydroxyl radical which is an important oxidizing species taking part in the photodegradation of RB dye. The holes generated in the valance band during solar illumination may also play an active role in the degradation of RB dye. Thus, the combination of hydroxyl radical and holes vibrantly take part in photodegradation of the RB dye. Liquid chromatography mass spectrometry (LC-MS) was used to confirm the www.nature.com/scientificreports www.nature.com/scientificreports/ possible degradation and intermediates formed after 80 min. Sample with minimum intensity in the UV-visible spectra was used for LC-MS analysis. Figure 8 shows the mass spectra of the RB dye solution after photocatalytic reaction, which shows that after 80 min no peak corresponding to RB dye is observed confirming the degradation of the dye. The different possible fragments obtained during the process of photocatalysis are presented in Fig. 9.
The kinetics of degradation process of RB dye was modelled with the Langmuir-Hinshelwood mechanism 71 . The importance of this model is that it also covers adsorptions properties of the photocatalyst. The apparent rate constant of degradation of RB was evaluated by using Langmuir-Hinshelwood model. From experimental data, the rate constant was calculated by using linear fit and was equal to 0.018/min for NaNbO 3 with R 2 equal to 0.98. Equation 1 is the kinetic equation used to determine the rate constant for the photocatalytic reaction.  Dielectric properties. Frequency and temperature dependent dielectric properties of NaNbO 3 pellets sintered at two different temperatures (500 °C and 700 °C) were measured from 20 Hz to 1 MHz at different temperatures ranging from 25 °C to 500 °C. Figure S6 shows the variation of dielectric constant (ɛ) and dielectric loss (D) of the NaNbO 3 samples sintered at 500 °C and 700 °C with frequency at room temperature. The variation of (ɛ) and (D) of NaNbO 3 nanoparticles sintered at 500 °C and 700 °C with frequency at different temperatures (50 °C to 500 °C) is shown in Figs 11 and 12, respectively. The dielectric constant (ɛ) is found almost independent on frequency at low temperature <250 °C, however, with increase in frequency there is an appreciable decrease in dielectric loss (D) values. With increase in temperature, the change in (ɛ) and (D) with respect to frequency becomes more prominent as shown in Figs 11 and 12. Both (ɛ) and (D) have high value at low frequencies which shows decreasing trend with increase in frequency. The decrease in (ɛ) and (D) is gradual at low frequencies and becomes almost constant at higher frequencies. This declined behaviour can be explained by long-time polarization effect caused by certain structures such as space charges, dipoles which do not have ability to contribute to the overall polarization at higher frequencies 72 . At low frequencies, space charge polarization does undergo the relaxation process which follows the frequency of the applied field easily whereas, at high frequencies, these space charges do not get enough time to orient themselves along the frequencies of applied field 73 . The reason for the decrease of (ɛ) and (D) with frequency at different temperatures is that electric dipoles are not capable to follow the applied electric field as explained by Maxwell-Wagner interfacial polarization 74 . The change of (D) with  www.nature.com/scientificreports www.nature.com/scientificreports/ respect to frequencies can be explained by Koop's phenomenological theory 75 . Low loss value can be attributed to the nano dimensional particles 76 .
The variation of (ɛ) and (D) with respect to temperature measured at different frequencies (1kHz-1MHz) is shown in Fig. 13a-d for samples sintered at 500 °C and 700 °C respectively. From the results obtained, it is observed that with increase in temperature both (ɛ) and (D) increases at different frequencies. At temperature above 300 °C, both dielectric loss as well as dielectric constant increases sharply which may be the result of increased conductivity. Figure 14a,b represents the variation of (ɛ) and (D) of samples sintered at 500 °C and  www.nature.com/scientificreports www.nature.com/scientificreports/ 700 °C with temperature at 500 kHz frequency respectively. The results obtained show that sample sintered at 500 °C show stable value of (ɛ) and low (D) upto 250 °C. Above 250 °C, both (ɛ) and (D) increases sharply, therefore, upto 250 °C NaNbO 3 nanoparticles sintered at 500 °C could be used as stable dielectric material in electronic devices. However, the sample sintered at 700 °C shows the same effect in the properties upto 200 °C.
The dielectric constant of the NaNbO 3 nanoparticles was found to be 41 and 38.5 for the samples sintered at 500 °C and 700 °C, respectively at 500 kHz. The increase in sintering temperature increases the crystallinity and hence the crystallite size of nanoparticles is increased. Therefore, the sample sintered at high temperature has fewer lattice defects compared to the sample having small crystallite size and low sintering temperature 77 . According to conductivity model, the ac conductivity of the materials is directly proportional to the amount of defect sites present in the material 77 . Therefore, at high sintering temperature the sample has minor defect sites and may have low conductivity values as discussed later in ac conductivity section. Dielectric constant and square root of conductivity of the materials are directly proportional to each other. From ac conductivity results, with increase in sintering temperature, the ac conductivity decreases therefore, dielectric constant of NaNbO 3 nanoparticles decreases. Figure S7 shows the TEM images of the NaNbO 3 nanoparticles sintered at two different temperatures. Figure S7a represents the TEM image of the sample sintered at 500 °C while as Fig. S7b shows the TEM micrograph of the sample sintered at 700 °C. The change in (ɛ) and (D) at different frequencies and temperatures are tabulated in Supplementary Tables 2S-5S. This change in (ɛ) and (D) with temperature of the samples sintered at different temperatures is plotted in Supplementary Fig. S8 which shows that with increase in temperature, the difference in the (ɛ) and (D) at initial (20 Hz) and final (1 MHz) frequency increases. AC conductivity. Figure 15a,b shows the dependence of ac conductivity on frequency of the samples sintered at 500 °C and 700 °C respectively. For the sample (sintered at 500 °C), upto temperature <250 °C there is negligible change in conductivity of the NaNbO 3 nanoparticles. However, above 250 °C temperature, the conductivity increases sharply throughout the frequency range. The increase in conductivity is ascribed to the increase in  www.nature.com/scientificreports www.nature.com/scientificreports/ the number of charge carriers due to formation of large oxygen vacancies in the lattice at high temperatures 78 . Similarly, for the sample (sintered at 700 °C), the conductivity almost remains independent on frequency upto 300 °C, however, above this temperature, the conductivity shows sharp increment with increase in frequency. Fitting of power law (σ = Aω η ) was used to understand the variation of conductivity of the samples with frequency at different temperatures. The value of "η" determines the mechanism of the conductivity. For η > 1, conductivity follows the Maxwell Wagner (M-W) relaxation process 79 , while as, for η < 1 conduction process follows correlated barrier hopping (CBH) mechanism 79 . By applying power law to the conductivity both the mechanisms were followed by NaNbO 3 nanoparticles at different temperatures. The NaNbO 3 nanoparticles sintered at different temperatures show η > 1 upto 350 °C, suggesting that samples follow M-W mechanism of conduction. However above 350 °C, η < 1 therefore CBH mechanism is responsible for the conduction. The conduction mechanism is clearly visible in the frequency range 20Hz-1MHz, where the η starts decreasing with increase in temperature suggesting that CBH process starts to dominate at high temperature and above 350 °C, CBH completely overcomes the M-W mechanism. This may be since at higher temperature the charge carriers get enough thermal energy to show the conduction barrier hopping process. The value of η at different temperatures is tabulated in Table 6S.

Conclusion
In summary, high surface area nanosized sodium niobate has been successfully fabricated at low temperature by facile polymeric citrate precursor route. The synthesized nanoparticles were characterized by using XRD, TEM, XPS and BET surface area analysis. The synthesized NaNbO 3 nanoparticles were directly employed as electrode material for OER and HER activity, which shows promising results like significant current densities, comparable onset potential and tafel slopes for OER and HER with respect to state of art electrocatalysts. Electrocatalytic properties of nanosized and bulk NaNbO 3 samples were carried out which demonstrated that nanosized NaNbO 3 show better activity as compared to bulk counterpart. Photocatalytic studies were carried out for over 80 min that showed enhanced catalytic degradation (~86%) of the organic dye (RB) on the surface of NaNbO 3 . The substantial electrocatalytic and photocatalytic performances were ascribed to the large surface area of NaNbO 3 catalyst. The change of dielectric properties with operating frequency, temperature and sintering temperature have been discussed considering Maxwell-Wagner and Koop's theory. It was observed that with increase in sintering temperature from 500 °C to 700 °C there is slight reduction of dielectric constant from 41 to 38.5 at 500 kHz frequency. Conductivity process of the samples was understood by applying power law fitting. The present study offers strategy for the design of new multifunctional materials for energy conversion techniques. Synthesis of Nano-sized NaNbO 3 . NaNbO 3 nanoparticles were synthesized at low temperature by using polymeric citrate precursor method. 0.1 M aqueous solution of sodium hydroxide was prepared in deionised water. Citric acid was added in 25 ml of aqueous sodium hydroxide solution and was stirred for about 10 min. To this solution, 0.025 moles of niobium chloride was added and stirred for 3 h followed by addition of ethylene glycol. Then, the solution mixture was stirred and heated at 55 ± 5 °C continuously until the formation of viscous gel. Then, the reaction mixture was placed inside the muffle furnace and temperature was increased up to 135 °C for 20 h. After 20 hrs, temperature was further raised and kept at 300 °C for 2 hrs. A black mass referred to as precursor was obtained. The as-obtained precursor was grinded and calcined in air at 500 °C for 12 h and was used for further characterization and studies. The molar ratio of Citric acid: metal ion: Ethylene glycol during the reaction was 40: 1: 10. synthesis of bulk NaNbo 3 . In a typical synthesis process, 10 N NaOH solution was prepared and was kept on stirring. To this solution 0.025 moles of Nb 2 O 5 was added and stirred for 1 h. The reaction mixture was then shifted to 50 ml teflon lined autoclave and was heated in muffle furnace at 150 °C for 4 h. The reaction mixture was then left to cool down naturally. Final product was collected by centrifugation and washing.

Materials
Characterization. The powder X-ray diffraction technique (PXRD) was used to determine the crystallinity and phase purity. Rikagu X-ray diffractometer equipped with Cu Kα radiation (λ = 1.5406 Å) was used to obtain the diffraction patterns at the normal scans with a step size of 0.5 °/s and a step time of 1 s with 2θ ranging from 20° to 70°. The XRD pattern and peak positions obtained were compared with the JCPDS standard files to identify the phase and lattice parameters of the synthesized material. To estimate the exact shape and size of the as-synthesized nanoparticles, Transmission electron microscopic (TEM) study was carried out by using FEI Technai G 2 20 HRTEM with an accelerating voltage of 200 kV. To exactly determine the chemical composition and binding energies of different components of the synthesized material, the X-ray photoelectron spectroscopic studies were carried out by using EAC200 SPHERA 547 having Mg Kα as radiation source. BET surface area analyser (Nova 2000e, Quantachrome Instruments Limited, USA) was employed to estimate the specific surface area and the pore size of the sample in the presence of liquid nitrogen (77 K). To remove the adsorbed gases and moisture from the sample, vacuum degassing was carried out for 3 h at 200 °C. Specific surface area was determined by using Multipoint BET equation and pore size distribution was determined by using Dubinin-Astakhov (DA) method. All the calculations made in the BET study are done automatically by the instrument using the software-based equations to compute specific surface area and pore size distributions. (2019) 9:4488 | https://doi.org/10.1038/s41598-019-40745-w www.nature.com/scientificreports www.nature.com/scientificreports/ electrocatalytic measurements. The electrochemical measurements were carried out with CHI 660E, China electrochemical analyser having a standard three electrode cell testing system at 30 °C. The synthesized sodium niobate nanoparticles were employed as working electrode while, Ag/AgCl and Pt wire were used as reference and counter electrode respectively. The working electrodes were prepared by putting a drop of sample slurry over the surface of glassy carbon electrode (GCE) and then dried prior to use it. The slurry was prepared by using 5.0 mg NaNbO 3 electro-catalyst with 0.025 ml Nafion and 1 ml propanol. The resulting mixture was then sonicated for 30 minutes to get the homogeneous mixture. The loaded mass of the NaNbO 3 electro-catalysts on the GCE was ~0.2 mg/cm 2 . The surface area of electrode was 0.07 cm 2 , which was used to calculate the current density of electrode materials. Cyclic voltammetry (CV) was used to determine hydrogen evolution reaction and oxygen evolution reaction activity of NaNbO 3 electrode at scan rate 100 mV/s in 0.5 M KOH electrolyte. CV measurements of sodium niobate nanoparticles for HER and OER were evaluated in a peak window range from 0 to −1.4 V. Linear sweep voltammetry curve (LSV) was obtained in cathodic as well as anodic directions and was used to calculate the onset potential and current density for HER and OER. The Nernst equation can be employed for the conversion of the potential of OER and HER versus Ag/AgCl electrode to RHE at room temperature i.e. Photocatalytic experiment. The activity of NaNbO 3 nanoparticles as photocatalyst was assessed by monitoring the degradation of (RB) as a probe reaction. Photocatalytic reaction was carried out in presence of sunlight at an atmospheric temperature (35 °C). 20 mg nanoparticles were dispersed in 50 ml RB dye solution having 1 × 10 −5 M concentration. The suspension obtained was stirred continuously in the dark conditions for 30 min to attain the adsorption desorption equilibrium between organic dye and the catalyst. After reaching adsorption desorption equilibrium, photocatalytic reaction was initiated by irradiation with sunlight. The spectrum was recorded at the interval of 10 min during the total period of 80 mins for analysis. Other experiments were also carried out either in absence of catalysts or under dark conditions to authorize that the degradation reaction is carried out by photocatalysis only. The process of photodegradation of RB dye was observed by measuring the characteristic change in absorption intensity at 545 nm using UV-visible spectrophotometer. The percentage photocatalytic degradation was calculated using following equation.
where C 0 is the RB dye concentration at t = 0 i,e., after attainment of adsorption-desorption equilibrium before irradiation and C is the concentration of dye after time interval t. Liquid chromatography mass spectroscopy (LC-MS) of the dye solution was carried out to attest whether the dye has been degraded or not.
Dielectric measurements. Dielectric measurements of the NaNbO 3 samples were carried out in air at temperatures ranging from 25-500 °C over a frequency range of 20Hz-1MHz. HF-LCR meter (6505 P, Wayne Kerr Electronics, UK) was used to carry out the dielectric measurements. For dielectric measurements, principle of parallel plate capacitor was used. These measurements were carried out by using 8 mm disk shaped pellet with a thickness of 0.5 mm prepared by applying uniaxial pressure of 5 tons and using 5% polyvinyl alcohol (PVA) as a binder. The pressure was applied using the KBr press Model M-5 (technosearch instruments). The prepared pellets were sintered at 500 °C and 700 °C temperatures to remove the binder from the sample. The pellets were coated with thin layer of silver paint (Ted Pella, Inc.) to form conducting contacts which act as electrode. Virtual instrument package LABVIEW (National Instruments) in interference with LCR meter was used to collect the data.