The Impact of Co Doping and Annealing Temperature on the Electrochemical Performance and Structural Characteristics of SnO2 Nanoparticulate Photoanodes

Obtaining H2 energy from H2O using the most abundant solar radiation is an outstanding approach to zero pollution. This work focuses on studying the effect of Co doping and calcination on the structure, morphology, and optical properties of spin-coated SnO2 films as well as their photoelectrochemical (PEC) efficiency. The structures and morphologies of the films were investigated by XRD, AFM, and Raman spectra. The results confirmed the preparation of SnO2 of the rutile phase, with crystallite sizes in the range of 18.4–29.2 nm. AFM showed the granular structure and smooth surfaces having limited roughness. UV-Vis spectroscopy showed that the absorption spectra depend on the calcination temperature and the Co content, and the films have optical bandgap (Eg) in the range of 3.67–3.93 eV. The prepared samples were applied for the PEC hydrogen generation after optimizing the sample doping ratio, using electrolyte (HCl, Na2SO4, NaOH), electrode reusability, applied temperature, and monochromatic illumination. Additionally, the electrode stability, thermodynamic parameters, conversion efficiency, number of hydrogen moles, and PEC impedance were evaluated and discussed, while the SnO2 films were used as working electrodes and platinum sheet as an auxiliary or counter electrode (2-electrode system) and both were dipped in the electrolyte. The highest photocurrent (21.25 mA/cm2), number of hydrogen moles (20.4 mmol/h.cm2), incident photon-to-current change efficiency (6.892%@307 nm and +1 V), and the absorbed photon-to-current conversion efficiency (4.61% at ~500 nm and +1 V) were recorded for the 2.5% Co-doped SnO2 photoanode that annealed at 673 K.


Introduction
Developing an efficient and clean energy technology with zero pollution became the centre of attention to meet the energy demand at the time of fossil fuel depletion. As an outstanding approach, converting renewable solar energy to highly efficient, eco-friendly, and affordable H 2 energy using H 2 O splitting is attracting intensive research interest worldwide [1,2]. In addition to CO 2 zero-emission, the H 2 energy value is~122-142 kJ/g which is 2.8-3.0 times larger than the energy value obtained from liquid hydrocarbons used as fuel [3,4]. In this context, many attempts have been carried out to adapt various transparent metal oxides (TCOs) as photoanode in the electrochemical cell for H 2 energy generation. Among them is a stannic oxide (SnO 2 ). SnO 2 is nontoxic, highly transparent, abundant in nature, has good electrical conductivity, and has a high electron mobility of~(1-2) × 10 2 cm 2 /V·s. Additionally, the photocatalytic activity toward RhB, when exposed to visible light, is increased, according to Toloman et al. [8].
The material's crystallinity is mainly relayed on the heat treatment and this crystallinity affects the E g , rate of electron-hole recombination, and the efficiency of energy conversion [21]. Based on the literature survey, there is no complete report on the effect of Co doping and heat treatment on structural properties, water splitting ability, and H 2 production efficiency of SnO 2 thin films. This work aims to demonstrate for the first time the effects of Co doping and calcination temperature (CT) on the structure, morphology, optical characteristics, and PEC catalytic activity of the spin-coated SnO 2 photoelectrodes. In addition to examining the impact of PEC temperature on the performance of the optimised photoelectrode, the conversion efficiencies under white and monochromatic light illumination and the number of hydrogen moles generated are also assessed. In particular, this study focuses on optimizing the CT and Co doping level to increase the optical absorption and surface area by enhancing the optical and textural aspects of PEC SnO 2 -based electrodes for solar H 2 generation.

Materials and Preparation
Un-doped and 2.5, 5.0, and 7.5 at.% Co-doped SnO 2 thin films were prepared by dissolving~1.13 g SnCl 2 .2H 2 O (Merck, Darmstadt, Germany) of molecular weight (M W ) = 225.63 g/mol in 10 mL pure ethanol. For the doped solution, certain amounts of Co(NO 3 ) 2 . 6H 2 O or CoH 12 N 2 O 12 (Aldrich, St. Louis, MO, USA) of M W = 291.03 were added. The solutions were stirred at 323 K for 2 h till getting clear and homogeneous solutions and aged for 24 h at (300 K). The glass slides of specific dimensions were treated by sonication in acetone, methanol, and deionised water before being air dried. The sol-gel produced solutions were deposited on the glass slides at 2000 rpm for 25 s, then dried at 473 K for 5 min on an electric hot plate. Six coatings were performed to increase the films' thickness, before the final calcination at 673, 773, and 873 K for 2 h in an air furnace. In each case, the furnace was left to cool to RT and this process takes 8-10 h. Before use, the samples were kept in zipper bags to avoid the moisture effect.

Measurements
Investigating phase purity and crystallinity was done by using X-ray diffraction (X'PertPro MRD, Philips, Malvern, UK) by assisting with Cu-K α radiation of wavelength 0.1542 nm during steps of 0.02 • . 2D and 3D images to study the film's morphology, roughness, and thickness of the samples were taken by the atomic force microscope of non-contact mode (XE-100E, Park, Suwon, Korea). The vibrational properties of the films (Raman spectra) were recorded by Thermo Fisher (Austin, TX, USA) DXR Raman spectrometer with an 8 mW power laser beam of 532 nm wavelength. The absorbance spectra in the range of 200-1350 nm, for determination of the bandgaps, were recorded by UV-vis-NIR 3700 double beam Shimadzu spectrophotometer (Maryland, MD, USA). All these measurements were carried out at RT. The PEC H 2 O splitting was measured by a 2400-Keithley SourceMeter in 100 mL of 0.5 M HCl, NaOH, and Na 2 SO 4 solutions at RT with a scanning rate of 1 mV/s. The nanocomposite electrode with a 1 cm 2 surface area was used as the photoanode, and the counter electrode of the same area was the Pt-electrode. i.e., for PEC measurements, both a working 1 cm 2 electrode and a counter 1 cm 2 Pt electrode are used (2-electrode System). The photocurrent density-voltage (J ph -V) behaviours were measured under a standard white illuminance (AM 1.5 G, 100 mW/cm 2 ) and evaluated with the use of a 400 W mercury xenon light source (Newport, MODEL: 66926-500HX-R07, Newport, UK) and utilizing the 2400-Keithley SourceMeter (Tektronix company, Beaverton, OR, USA). Figure S1 (Supplementary Materials) shows XRD patterns of the as-deposited films (before annealing). The wide peak (halo) extends from 17 • to 38 • on the 2θ scale indicating the amorphous structure of the film. After doping with 2.5-7.5% Co, two peaks around 26 • and 34 • appear to be superimposed on this halo peak. This indicates that Co doping enhanced the crystallinity of the as-deposited films. This can take place when the Co 2+ ions form stable solid solutions with SnO 2 and take the regular lattice positions in the SnO 2 lattice [22].

XRD and AFM Study
XRD of the pure and Co-doped films calcined at temperatures in the range of 673-873 K are depicted in Figure 1a- [2]. XRD of spray deposited Pd-doped SnO 2 contained only one sharp (110) peak [18]. However, increasing the annealing temperature to 873 K encouraged the formation of the unstable SnO phase, indicated by a sign (*) and a peak at 31.85 • [23]. Similarly, SnO diffraction peaks were observed among those of SnO 2 that were prepared by the chemical vapour deposition at 600 • C [24].  Figure S1 (Supplementary Materials) shows XRD patterns of the as-deposited films (before annealing). The wide peak (halo) extends from 17° to 38° on the 2θ scale indicating the amorphous structure of the film. After doping with 2.5-7.5% Co, two peaks around 26° and 34° appear to be superimposed on this halo peak. This indicates that Co doping enhanced the crystallinity of the as-deposited films. This can take place when the Co 2+ ions form stable solid solutions with SnO2 and take the regular lattice positions in the SnO2 lattice [22].

XRD and AFM Study
XRD of the pure and Co-doped films calcined at temperatures in the range of 673-873 K are depicted in Figure 1a-c. As seen, all films show polycrystalline structures. The observed diffraction peaks at ~ 26.63°, 33.87°, 37.92°, 51.78°, and 64.72° correspond to (110), (101), (210), (211) and (112) planes of SnO2 tetragonal phase, respectively. These wide peaks indicate the formation of nano-crystalline SnO2 materials. No peaks related to Co metal or one of its oxides are detected. A similar result was reported for the chemically precipitated Fe-doped SnO2 NPs [2]. XRD of spray deposited Pd-doped SnO2 contained only one sharp (110) peak [18]. However, increasing the annealing temperature to 873 K encouraged the formation of the unstable SnO phase, indicated by a sign (*) and a peak at 31.85° [23]. Similarly, SnO diffraction peaks were observed among those of SnO2 that were prepared by the chemical vapour deposition at 600 °C [24].  The preferred orientation feature can be quantified by the texture coefficient (TC) that signifies the texture of a specific plane. The thin film with randomly oriented crystallites presents TC = 1, and TC > 1 implies the preferred growth, i.e., a larger abundance of crystallites oriented in a (hkl) direction. TC was calculated considering the main three peaks; (110), (101), and (211), using XRD data and the following equation: [25], where I(hkl) and Io(hkl) refer to the relative intensity and the standard JCPDS intensity of the plane (hkl), respectively, and Z is the reflections. From Table 1, it is noted that TC (110) and TC (101) are larger than unity. This suggests the existence of more crystals related to these two peaks. The origins of the noted orientation of the (110) plane can be explained by the periodic band chains theory (PBC) [26]. Previously, it was reported that doping with more than 3.0 wt.% Er in the e-beam evaporated snO 2 films changed the preferred growth direction from (110) to (302) [19]. The CT (673-873 K) does not affect the peaks' position. Additionally, it is seen from Figure 1 that the doping with 2.5-7.5 at.% Co causes no peaks' shift. This may be because the two ions Co 2+ and Sn 4+ have very close ionic radii of 0.74 Å and 0.71 Å, respectively [20]. Carbon doping (10-50 wt.%) into SnO 2 resulted in a slight shift (by 0.26 • ) towards lower 2θ values [3]. Figure 1a demonstrates that for the undoped film, the calcinated sample at 773 K had more intense XRD peaks than the calcinated sample at 673 K. This suggests that the film's crystallinity improved when the CT was raised from 673 to 773 K. A similar observation was reported for SnO 2 NPs prepared by continuous flow hydrothermal route, at pH = 3 and annealing temperatures in the range of 523-673 K [7]. However, increasing the temperature to 873 K made the peaks of reduced intensity. This may ascribe to the mismatching between the film and the glass substrates which may be bent at this high temperature. SnO 2 film doped with 2.5 at.% Co and calcined at 673 or 773 K exhibits peaks of relatively higher intensity compared with those of the un-doped film indicating an additional increase in the film's crystallinity. Increasing Co content to 5.0 and 7.5 at.% reduces the peaks' intensity of the film calcined at 673 K, but this intensity increases with increasing temperature to 773 and 873 K. The broadness of SnO 2 diffraction lines with increasing Co content suggests the intercalation of Co 2+ into SnO 2 and these ions create specific disorders and lead to compositional heterogeneity [2]. The crystallite size (C s ) and the dislocation density (δ = 1 C s 2 ) of the samples were calculated utilizing the Scherer equation; C s = 0.9λ/Γ cos θ, where λ = 0.154 nm (the wavelength emitted from the K α -Cu source), Γ is the full-width at half maximum intensity, considering the main three peaks; (110), (101) and (211). The obtained values are listed in Table 1. As can be noticed, Co doping and CT have a significant effect on both C s and δ. All films have C s in the range of 18.4-29.2 nm. Higher content of Co (5.0-7.5 at.%) leads to a decrease in C s , however, increasing CT hinders this increase. Similarly, excessive doping with Ta (2.6-15.5 at.%) resulted in deterioration of crystallinity of SnO 2 film prepared by radio-frequency sputtering [9]. Er-doping significantly increased the crystallite size of SnO 2 films from 17.3 nm to~29.5 nm, which was then marginally changed with increasing Er content [19]. The obtained results mean that the films' crystallinity can be controlled by Co doping and calcination.
The films' surface morphology was examined by AFM via capturing 2D and 3D images, Figure 2a-f and Figure S2 (Supplementary Materials). As seen from Figure S2, the unannealed pure film exhibits no definite morphology, however, the 5.0% Co doped film exhibits a nanoparticulate morphology with a relatively high particle density per unit area. The 3D and 2D images of 2.5 at.%-doped SnO 2 calcined at 673 K, Figure 3a,b illustrates that the surface is of granular morphology, i.e., the surface is composed of a very large number of spherical-shaped particles and size less than 20 nm and average root-mean-square roughness (R rms ) of~18 nm. Increasing the CT to 773 K led to a larger grain size of 23 nm and reduced R rms to 14 nm, Figure 2c,d. However, increasing Co content to 5.0 at.% decreased the grain size to 22 nm and the R rms to 11 nm, Figure 2e,f. The obtained thicknesses of the films were ranged from 255 to 270 nm, as mentioned in Table 2. Thus, the prepared films are of relatively smooth surfaces and are useful for optoelectronic and photocatalytic applications [9]. Figure 2g-i shows the role of CT (673-873 K) on the SnO 2 films doped with the higher content of Co (7.5 at.%). As seen, both the grain size and surface roughness increase with annealing temperature. Similarly, the chemically precipitated Fe-SnO 2 appeared under FE-SEM in the form of blocks of 500 nm in size that were composed of NPs with an average size of ≈25 nm [2]. These findings show that the measurements made by XRD and AFM are consistent.

Raman Spectral Analysis and Optical Properties
For deeper insights and further confirmation of the SnO 2 phase under calcination and Co loadings, Raman spectra were recorded and presented in Figure 3. Four peaks appear in the spectra of the films; two weak peaks at 452 and 780 cm −1 and two strong wide peaks at 552 and 1090 cm −1 . The 452 cm −1 peak (E g mode) and 780 cm −1 peak (B 2g mode) are due to the vibration of O in the O plane and the construction/expansion of Sn-O bonds, respectively, which usually occur in nanocrystalline tin oxide [25]. B 2g mode is slightly shifted under the effect of CT and Co loading. The broad peak at 552 cm −1 is related to the rutile phase of SnO 2 [19]. This result is consistent with XRD results. As seen, increasing the CT and Co content resulted in changing the intensity of this peak. The most intense broad peak at 1090 cm −1 is related to the multi-phonon transitions [27]. The E g mode is highly sensitive to oxygen vacancies, and its appearance indicates that our samples can be used for photocatalytic applications and photoelectrochemical cells.

Raman Spectral Analysis and Optical Properties
For deeper insights and further confirmation of the SnO2 phase under calcination and Co loadings, Raman spectra were recorded and presented in Figure 3. Four peaks appear in the spectra of the films; two weak peaks at 452 and 780 cm −1 and two strong wide peaks at 552 and 1090 cm −1 . The 452 cm −1 peak (Eg mode) and 780 cm −1 peak (B2g mode) are due to the vibration of O in the O plane and the construction/expansion of Sn-O bonds, respectively, which usually occur in nanocrystalline tin oxide [25]. B2g mode is slightly shifted under the effect of CT and Co loading. The broad peak at 552 cm −1 is related to the rutile phase of SnO2 [19]. This result is consistent with XRD results. As seen, increasing the CT and Co content resulted in changing the intensity of this peak. The most intense broad peak at 1090 cm −1 is related to the multi-phonon transitions [27]. The Eg mode is highly sensitive to oxygen vacancies, and its appearance indicates that our samples can be used for photocatalytic applications and photoelectrochemical cells.    [28], SnO 2 exhibits a substantial UV absorbance band (270-300 nm). The clear interference fringes that are seen in all films confirm the smoothness and homogeneity of the films seen in AFM images. The 2.5% Co-doped film annealed at 673 K shows relatively higher absorption in the studied wavelength range. The interference fringes of films calcined at 773 K begin at a lower wavelength and their number is larger than the films calcined at 673 K. Increasing CT resulted in denser films of more compacted layers. The direct optical band gap E g can be evaluated utilizing Tauc's formula (αhυ) 2 = A(hυ-E g ), whereas the incident photon energy hυ (eV) = 1242 λ (nm) , A is a constant, and α is the absorption coefficient (α = 2.303 Absorption/film thickness). The insets of Figure 4a,b show the (αhυ) 2 vs. hυ plots. E g is obtained by extrapolating the linear portion to zero absorption and Table 2 provides their values. E g values are between 3.67 and 3.93 eV. The decrease in E g of SnO 2 with doping till 5.0 at.% is related to the introduction of electronic states in its bandgap. Although it has been reported that increasing CT increases oxygen vacancies [29], the effect of enhanced crystallisation outweighs the effect of oxygen vacancy concentration (Ov), which results in a decrease in the number of charge carriers inside the material in films that are calcined at 773 K. The drop in E g from 3.73 to 3.67 eV at a doping level of 7.5 at.% may be related to an elevated Ov. This is also anticipated at CT = 873 K, where the presence of the low valence state (Sn 2+ ), depicted in Figure 1c, results in an increase in Ov and unbound electrons that lower the conduction band's bottom and, consequently, reduce E g . In rare circumstances, the Ov can introduce mid-gap states by acting as shallow donors [30]. 673 K. Increasing CT resulted in denser films of more compacted layers. The direct optical band gap Eg can be evaluated utilizing Tauc's formula (αhυ) 2 = A(hυ-Eg), whereas the incident photon energy ℎ eV , A is a constant, and α is the absorption coefficient (α = 2.303 Absorption/film thickness). The insets of Figure 4a,b show the (αhυ) 2 vs. hυ plots. Eg is obtained by extrapolating the linear portion to zero absorption and Table 2 provides their values. Eg values are between 3.67 and 3.93 eV. The decrease in Eg of SnO2 with doping till 5.0 at.% is related to the introduction of electronic states in its bandgap. Although it has been reported that increasing CT increases oxygen vacancies [29], the effect of enhanced crystallisation outweighs the effect of oxygen vacancy concentration (Ov), which results in a decrease in the number of charge carriers inside the material in films that are calcined at 773 K. The drop in Eg from 3.73 to 3.67 eV at a doping level of 7.5 at.% may be related to an elevated Ov. This is also anticipated at CT = 873 K, where the presence of the low valence state (Sn 2+ ), depicted in Figure 1c, results in an increase in Ov and unbound electrons that lower the conduction band's bottom and, consequently, reduce Eg. In rare circumstances, the Ov can introduce mid-gap states by acting as shallow donors [30]. Therefore, Co doping and calculations can be effectively used to tune the optical properties of SnO2. Yang et al. [17] reported a decrease in the Eg of SnO2 from 3.45 to 3.33 eV after doping with 7.0% (Cu and N). Then, 18 at.% Mo doping into SnO2 prepared by the magnetron co-sputtering technique shrank the Eg from 3.65 to 2.7 eV [14]. Additionally, 10 at.% (Co, Ni) doping reduced the Eg of SnO2 NPs synthesized using a green process from 3.33 to 2.08 eV [20]. Moreover, Pd doping (1-5 wt.%) narrowed Eg of SnO2 from 3.65 to 3.25 eV [18]. Ta doping in the range of 2.6-7.8 at.%) widen Eg of SnO2 from 3.99 to 4.21 eV that then reduced it to 4.17 eV at 15.5 at.% Ta. In addition, Er doping (0-3 wt.%) for the e-beam evaporated SnO2 film, 100 nm thickness, narrowed its Eg from 3.88 to 3.63 eV, then it widen to 3.72 Therefore, Co doping and calculations can be effectively used to tune the optical properties of SnO 2 . Yang et al. [17] reported a decrease in the E g of SnO 2 from 3.45 to 3.33 eV after doping with 7.0% (Cu and N). Then, 18 at.% Mo doping into SnO 2 prepared by the magnetron co-sputtering technique shrank the E g from 3.65 to 2.7 eV [14]. Additionally, 10 at.% (Co, Ni) doping reduced the E g of SnO 2 NPs synthesized using a green process from 3.33 to 2.08 eV [20]. Moreover, Pd doping (1-5 wt.%) narrowed E g of SnO 2 from 3.65 to 3.25 eV [18]. Ta doping in the range of (2.6-7.8 at.%) widen E g of SnO 2 from 3.99 to 4.21 eV that then reduced it to 4.17 eV at 15.5 at.% Ta. In addition, Er doping (0-3 wt.%) for the e-beam evaporated SnO 2 film, 100 nm thickness, narrowed its E g from 3.88 to 3.63 eV, then it widen to 3.72 eV at 5.0 wt.% Er [19]. Thus, doping with transition metal is a facial method to reduce E g of SnO 2 to be more valid for photocatalytic application under visible light exposure.

Influence of the Doping Ratio and the Used Electrolyte
The PEC cell electrolyte is one of the essential things that has a major impact on the PEC performance of the photoelectrode. The photocatalytic interfaces will differ in several ways depending on how the electrolytes are changed. The chosen electrolyte must be compatible with the PEC catalyst being used. Electrochemical interactions between photo-electrodes and electrolytes are prohibited, and their optical absorption bands must not overlap. The electrolyte structure, which alters the conductivity of electrolytes and affects ion transport, has a substantial impact on the rate of PEC H 2 generation. The rate of H 2 oxidation and the catalytic effectiveness of the PEC are influenced by the electrolyte's conductivity and pH [31]. The electrolyte impedance may decrease due to insufficient active cations or anions like Na+/SO 4 2− or OH − , which may stimulate band bending. To pinpoint the main differences in the PEC performances of the examined photoanodes; a strong acid (HCl, pH = 0.3), a strong basic (NaOH, pH = 13.69), and a neutral salt (Na 2 SO 4 ) were all used at the concentration (0.5 M). In Figure 5b, the current density rises in the following order: J ph (NaOH) > J ph (HCL) > J ph (Na 2 SO 4 ), with the J ph 's value growing as the utilised bias rises. The maximum J ph value in solutions of NaOH, HCL, and Na 2 SO 4 is 21.25 mA/cm 2 , 10.972 mA/cm 2 , and 5.09 mA/cm 2 , respectively. Therefore, NaOH is the ideal electrolyte for the PEC process using 2.5% Co-doped SnO 2 anode calcination temperature at 673 K, abbreviated to (2.5% Co, 673 K) SnO 2 . By reducing the Fermi level to a lower energy level through adsorbing extra H+/OH − species following the relationship E F = E F (pH 0 ) − 0:059 × pH (at RT), the surface charges of doped metal oxides and the separation of electrons from holes are both influenced by pH variations [32]. In other words, depending on the pH of the electrolyte solution, the fermi level for metal oxide semiconductors is located at various energy sites. The location of the Fermi level at the electrode's surface has an impact on how much the potential of the photo-generated electrons is reduced. The overpotential or motive factor for hydrogen synthesis reduces as E F 's energy level rises [33]. The J ph 's value for the (2.5% Co, 673 K), SnO 2 photoelectrode (21.25 mA/cm 2 ) is significantly high than that for the (0% Co, 673 K), and (7.5% Co, 673 K), SnO 2 photoelectrodes (8.43, 5.23, and 3.325 mA/cm 2 , respectively) under light exposure. despite the very slight difference in energy gap between them. Due to the surface's composition of many spherical particles with sizes less than 20 nm and higher R rms (18 nm) than other samples, it exhibits more absorption in the wavelength range under study and has a larger surface area. In order to get the highest photocurrent density in the NaOH electrolyte, the optimum photoanode is (2.5% Co, 673 K) SnO 2 . With the addition of OH − ions in that pH range, the measured current density in the various electrolytes increases at pH ≥ 7.
ways depending on how the electrolytes are changed. The chosen electrolyte must be compatible with the PEC catalyst being used. Electrochemical interactions between photoelectrodes and electrolytes are prohibited, and their optical absorption bands must not overlap. The electrolyte structure, which alters the conductivity of electrolytes and affects ion transport, has a substantial impact on the rate of PEC H2 generation. The rate of H2 oxidation and the catalytic effectiveness of the PEC are influenced by the electrolyteʹs conductivity and pH [31]. The electrolyte impedance may decrease due to insufficient active cations or anions like Na+/SO4 2− or OH−, which may stimulate band bending. To pinpoint the main differences in the PEC performances of the examined photoanodes; a strong acid (HCl, pH = 0.3), a strong basic (NaOH, pH = 13.69), and a neutral salt (Na2SO4) were all used at the concentration (0.5 M). In Figure 5b, the current density rises in the following order: Jph(NaOH) > Jph(HCL) > Jph(Na2SO4), with the Jph's value growing as the utilised bias rises. The maximum Jph value in solutions of NaOH, HCL, and Na2SO4 is 21.25 mA/cm 2 , 10.972 mA/cm 2 , and 5.09 mA/cm 2 , respectively. Therefore, NaOH is the ideal electrolyte for the PEC process using 2.5% Co-doped SnO2 anode calcination temperature at 673 K, abbreviated to (2.5% Co, 673 K) SnO2. By reducing the Fermi level to a lower energy level through adsorbing extra H+/OH− species following the relationship EF = EF(pH0) − 0:059 × pH (at RT), the surface charges of doped metal oxides and the separation of electrons from holes are both influenced by pH variations [32]. In other words, depending on the pH of the electrolyte solution, the fermi level for metal oxide semiconductors is located at various energy sites. The location of the Fermi level at the electrode's surface has an impact on how much the potential of the photo-generated electrons is reduced. The overpotential or motive factor for hydrogen synthesis reduces as EF's energy level rises [33]. The Jph's value for the (2.5% Co, 673 K), SnO2 photoelectrode (21.25 mA/cm 2 ) is significantly high than that for the (0% Co, 673 K), and (7.5% Co, 673 K), SnO2 photoelectrodes (8.43, 5.23, and 3.325 mA/cm 2 , respectively) under light exposure. despite the very slight difference in energy gap between them. Due to the surface's composition of many spherical particles with sizes less than 20 nm and higher Rrms (18 nm) than other samples, it exhibits more absorption in the wavelength range under study and has a larger surface area. In order to get the highest photocurrent density in the NaOH electrolyte, the optimum photoanode is (2.5% Co, 673 K) SnO2. With the addition of OH − ions in that pH range, the measured current density in the various electrolytes increases at pH ≥ 7.

PEC Reusability and Stability of the (2.5% Co, 673 K) SnO 2 Photoanode
In 0.5 M NaOH and utilizing a 2400-Keithley SourceMeter the reproducibility of the (2.5% Co, 673 K) SnO 2 photoanode was investigated by measuring the photocurrent density at +1 V in a standard white light environment for 12 consecutive runs, as illustrated in Figure 6a. The photocurrent density of the (2.5% Co, 673 K) SnO 2 photoanode was 19.28 mA/cm 2 in the first run, indicating that the photoelectrochemical water splitting mechanism is more efficient in light, as after 12 consecutive cycles, the photocurrent density decreased by 32.76%, from 19.28 to 12.99 mA/cm 2 , showing excellent reproducibility and reusability for the optimised SnO 2 photoanode in white light at RT. splitting cell, is defined as the proportion of overall hydrogen energy production to overall solar energy (AM 1.5 G, 100 mWcm 2 ). Equation (2) where Ptotal is the illuminating light's total power density in mW/cm 2 , A is the lighted section of the photoelectrode's area in cm 2 , and H2/S is the H2 molesʹ generation rate in mmol/sec. The computed STH for the (2.5% Co, 673 K) photoelectrode is 24.65%.  The PEC stability of the (2.5% Co, 673 K) photoanode is studied for a lengthy time in 0.5 M NaOH under the influence of white light and @+1 V between the photoanode and the platinum electrode. The change in photocurrent density with time is used to calculate the number of PEC H 2 moles as shown in Figure 6b [34]. Equation (1) was theoretically used to calculate the number of H 2 moles using Faraday's law [35]; where F = 9.65 × 10 4 C/mol (Faraday constant) and t is the elapsed time during the process. The number of hydrogen moles produced over time is presented in Figure 6b. This figure almost showed linear production of hydrogen, which can be used to confirm the stability of the designed PEC electrode. The obtained number of H 2 moles per active area theoretically is 41.4 mmol h −1 cm −2 . To measure the flow rate of the H 2 gas in SCCM, the outlet of the closed PEC unit is connected to the ALICAT Mass Flow Meter (Model: MB, 0-100SCCM (Standard mL/min)). Experimentally, the value is found to be 14.71 SCCM for our 1 cm −2 film which is equivalent to 39.45 mmol.h −1 .cm −2 . This value, which is pretty close to the value obtained theoretically, confirms the excellent effectiveness of the proposed photoelectrode.
The solar-to-hydrogen efficiency (STH), which is the overall efficiency of the PEC water splitting cell, is defined as the proportion of overall hydrogen energy production to overall solar energy (AM 1.5 G, 100 mWcm 2 ). Equation (2) was used to calculate STH % [36]: where P total is the illuminating light's total power density in mW/cm 2 , A is the lighted section of the photoelectrode's area in cm 2 , and H 2 /S is the H 2 moles' generation rate in mmol/sec. The computed STH for the (2.5% Co, 673 K) photoelectrode is 24.65%. Figure 7a shows how the PEC J ph -voltage curve of the (2.5% Co, 673 K) photoelectrodes in 0.5M NaOH electrolyte was affected by heating from RT to 363 K. The current density increased from 15.4 mA.cm −2 to 17.1 mA.cm −2 by rising T to 318 K and then reduced from 15.4 mA.cm −2 to 8.3 mA.cm −2 by raising T to 363 K, as can be seen in this figure. Evaluation of thermodynamic parameters including activation energy (E a ), enthalpy (∆H*), and entropy (∆s*) is also required. Figure 7b illustrates the relationship between the absolute temperature reciprocal (1/T) and the J ph (rate of reaction) of the photoelectrode of (2.5% Co, 673 K). The slope of the linear fitting in Figure 7b is used to calculate the value of E a using the Arrhenius equation [37]:

Effect of Temperature and Thermodynamic Parameters
where R is the universal gas constant (8.314 J/K.mol). From Figure 7b, slope = −E a /R and the E a of the (2.5% Co, 673 K) electrode is 10.032 kJ/mol. In addition, the values of ∆H* and ∆S* for the PEC H 2 production are estimated by using Eyring equation by drawing a relationship between Ln (J ph .T −1 ) and (T −1 ) as displayed in Figure 7c. The Eyring equation can be written as [38]: where K B = 1.38 × 10 −23 J/K (the Boltzmann's constant), and h = 6.626 × 10 −34 J.s (the Planck's constant). For (2.5% Co, 673 K), the linear fitting's slope gives a value of 12.78 kJ/mol for ∆H*, while its intercept gives a value of 311.113 J/mol for ∆S*. Figure 7a shows how the PEC Jph-voltage curve of the (2.5% Co, 673 K) photoelectrodes in 0.5M NaOH electrolyte was affected by heating from RT to 363 K. The current density increased from 15.4 mA.cm −2 to 17.1 mA.cm −2 by rising T to 318 K and then reduced from 15.4 mA.cm −2 to 8.3 mA.cm −2 by raising T to 363 K, as can be seen in this figure. Evaluation of thermodynamic parameters including activation energy (Ea), enthalpy (∆H*), and entropy (∆s*) is also required. Figure 7b illustrates the relationship between the absolute temperature reciprocal (1/T) and the Jph (rate of reaction) of the photoelectrode of (2.5% Co, 673 K). The slope of the linear fitting in Figure 7b is used to calculate the value of Ea using the Arrhenius equation [37]:

Effect of Temperature and Thermodynamic Parameters
where R is the universal gas constant (8.314 J/K. mol). From Figure 7b, slope = −Ea/R and the Ea of the (2.5% Co, 673 K) electrode is 10.032 kJ/mol. In addition, the values of ΔH* and ΔS* for the PEC H2 production are estimated by using Eyring equation by drawing a relationship between Ln (Jph.T −1 ) and (T −1 ) as displayed in Figure 7c. The Eyring equation can be written as [38]: where KB = 1.

Effect of Monochromatic Light Illumination and Conversion Efficiencies
The J ph is shown vs. applied voltage for the (2.5% Co, 673 K) photoanode in 0.5M NaOH at room temperature under monochromatic light in Figure 8a. A series of optical bandpass filters covering the visible spectrum from 307 to 636 nm were used. The maximum photocurrent, J ph = 17.06 mA.cm −2 , was obtained at 307 nm. The lowest photocurrent, J ph = 14.14 mA.cm −2 , was recorded at 636 nm as illustrated in the inset of Figure 8a. This current density-wavelength dependency may be related to the photo electrocatalytic response of the optimum photoelectrode for the H 2 production process and supports the absorption characteristic of the (2.5% Co, 673 K) photoelectrode in every wavelength. In general, this behaviour supports the photoelectrode's comprehensive response and ability to absorb a significant amount of the solar spectrum in the UV-Vis range.
photoelectrodeʹs maximal absorption. Optically related losses such as photon transmit-tance (Tr) and reflectance (R) were not taken into account in the IPCE calculation. To account for optical losses, it is required to assess the internal quantum efficiency, also referred to as the absorbed photon-to-current conversion efficiency (APCE%). The percentage of PEC-produced charges that contributed to the resulting photocurrent per absorbed photon is known as the APCE %.APCE is calculated by equation (6) [8,41]: where A(λ) is the optical absorbance. Figure 8c demonstrates the change of APCE% versus the incident λ. This figure shows three prominent APCE% values; 4.61% at ~ 500 nm, 4.595% around 470 nm, and 4.575% around 460 nm.  Measuring the incident photon-to-current conversion efficiency (IPCE) at various wavelengths further ensures the improved solar absorption of the (2.5% Co, 673 K) photoelectrode and its use for effective H 2 generation from H 2 O splitting. Equation (5) is used to get the IPCE% at a constant applied voltage (+1 V) [39,40]: where λ(nm) is the wavelength of the incident photons and P is the incident light power density of the Xenon lamp (Newport, 66142-500HX-R07) at each wavelength. Figure 8b represents the fluctuation of IPCE percentage with wavelength (λ). The (2.5% Co, 673 K) photoelectrode's highest IPCE % is 6.892% @307 nm, which is in line with the optimum photoelectrode's maximal absorption. Optically related losses such as photon transmittance (Tr) and reflectance (R) were not taken into account in the IPCE calculation. To account for optical losses, it is required to assess the internal quantum efficiency, also referred to as the absorbed photon-to-current conversion efficiency (APCE%). The percentage of PEC-produced charges that contributed to the resulting photocurrent per absorbed photon is known as the APCE%. APCE is calculated by Equation (6) [8,41]: where A(λ) is the optical absorbance. Figure 8c demonstrates the change of APCE% versus the incident λ. This figure shows three prominent APCE% values; 4.61% at~500 nm, 4.595% around 470 nm, and 4.575% around 460 nm. The extra electrical energy to the system must be removed when introducing a tiny applied voltage to the PEC system to estimate the photoelectrode catalytic efficacy. The applied bias photon-to-current efficiency (ABPE%) might be employed for this purpose. Equation (7) is used to calculate the ABPE% values for the designed photoelectrode [42,43]: where 1.23 is the standard state reversible potential of H 2 O and the externally applied voltage is denoted by Vapp. Figure 8d shows the ABPE % values for various electrolytes as a function of applied voltage. The highest value of ABPE% is 1.827%@NaOH 0.495 V, ABPE % of 1.07% @HCL 0.619 V and 0.19% @Na 2 SO 4 0.94 V, respectively. It is worth noting that the photoelectrode's good performance at low voltage might be beneficial for PEC cell setup.

PEC Impedance Spectroscopy (PEC-IS)
The charge transfer between the active photoelectrode and the electrolyte junction determines the photoelectrochemical system's impedance. PEC-IS study was made at room temperature using CHI electrochemical workstation (CH Instruments CHI660E) to investigate the dynamics of the charge carriers for the optimum (2.5% Co, 673 K) photoelectrode. The PEC-IS measurements' frequency (f ) range was 0.01-100,000 Hz at 0 V (vs. Ag/AgCl) under white light exposure. Figure 9a displays Nyquist plots of electrodes submerged in 0.5M NaOH at (0.0% Co, 673 K), (2.5% Co, 673 K), (5.0% Co, 673 K), and (7.5% Co, 673 K). The Bode charts in Figure 9b,c show the outcomes for the optimum (2.5% Co, 673 K) electrode. The redox reaction is shown by a narrow semicircle of a modest diameter on the high-f range of the (2.5% Co, 673 K) photoelectrode Nyquist plot [44], which is subsequently a linear stage filling the area of middle/low f s. The middle/low f s area is covered by a linear transition that follows the redox process [44]. These PEC-IS plots demonstrated mixed diffusion/kinetic controlled pathways. To further comprehend the PEC-IS observations via hydrogen evolution reaction (HER), the PEC-IS spectra were fitted to a straightforward analogous Randle circuit in the inset of Figure 9a. Using the ZSimpWin program, the Randle equivalent circuit was applied to simulate the PEC-IS spectra, as illustrated in Figure 9a. From Nyquist plot intercepts, the electrolyte resistance (R s = 171 Ω) is obtained at high f s. The semicircle diameter in the Nyquist plot is the same as the charge transfer resistance (R ct = 145 Ω). Concurrently, in the analogous circuit, the Warburg impedance is W = 0.0092 and the double-layer capacitance is C dl = 1.99 mF. The fundamental regulator of the HER is thus the charge transfer process (CTP), which is ensured by the uni-loop of the Nyquist plot. The optimum electrode can then generate a significant quantity of H 2 . The electron/hole recombination is typically deployed in conjunction with the charge transfer procedure (CTP) to regulate the HER. In other words, the small value of R ct that was observed indicates that the amount of charge recombination at the electrode/electrolyte interface was much decreased, which is consistent with the enhanced HER [45]. Bode charts for the (2.5% Co, 673 K) photoelectrode at the H 2 generation possibility in 0.5 M NaOH electrolyte at 25 • C are displayed in Figure 9b,c. Figure 9b depicts the overall impedance variation with f, and Figure 9c shows the phase variation with f. This chart illustrates capacitive contribution in between, along with resistive regimes at low and way higher f s. The double-layer capacitance (C dl ) of the photoelectrode and the charge transfer resistance (R ct ) relate to the low-f regime. The creation of a partially protective layer at the electrode's surface may be connected to the very high-f regime. At 16.63 Hz, the highest phase shift (θ max = 5.11 • ) is discovered. Figure 9c was used to estimate the charge carrier lifetime using the relation τ n = 1/2π ƒ max [46]. The obtained lifetime for the optimized photoelectrode is 9.57 ms. In addition, the data derived from Figure 9 show a significant decrease in charge recombination at the electrolyte/electrode interface. Along with enhanced ionic conductivity and electrolyte diffusion in the (2.5% Co, 673 K) photoelectrode, this also relates to a kinetically simple PEC system. Finally, the PEC performance was obtained in this study compared with many previously studied SnO 2 -based PEC electrodes as revealed in Table 3 [47][48][49][50][51][52][53][54][55][56][57][58]. The values of J, IPCE, and ABPE confirm that the (2.5% Co, 673 K) SnO 2 is an efficient electrode for the PEC H 2 O splitting under visible light exposure. Therefore, it has been concluded that the (2.5% Co, 673 K) SnO 2 electrode is very suitable for the PEC reactor. Finally, the PEC performance was obtained in this study compared with many previously studied SnO2-based PEC electrodes as revealed in Table 3 [47][48][49][50][51][52][53][54][55][56][57][58]. The values of J, IPCE, and ABPE confirm that the (2.5% Co, 673 K) SnO2 is an efficient electrode for the PEC H2O splitting under visible light exposure. Therefore, it has been concluded that the (2.5% Co, 673 K) SnO2 electrode is very suitable for the PEC reactor.

Conclusions
The influence of Co doping (0.0-7.5 at.%) and calcination temperatures in the range of (673-873K) on the structure, morphology, optical characteristics, and PEC performance of the sol-gel spin-coated SnO2 thin films photoanodes have been studied. The production of the SnO2 rutile phase with crystallite sizes in the range of 18.4-29.2 nm was validated

Conclusions
The influence of Co doping (0.0-7.5 at.%) and calcination temperatures in the range of (673-873 K) on the structure, morphology, optical characteristics, and PEC performance of the sol-gel spin-coated SnO 2 thin films photoanodes have been studied. The production of the SnO 2 rutile phase with crystallite sizes in the range of 18.4-29.2 nm was validated by XRD and Raman spectra. The granular structure and smooth surfaces with minimal roughness were revealed by AFM. UV-vis spectroscopy showed that the absorption of the films depends on Co content and annealing temperature, 2.5 at.% Co-doped film calcined at 673 K showed a higher absorption. Additionally, the optical bandgap (E g ) of all films was in the range of 3.67-3.93 eV and significantly depend on Co content and the calcination temperature. After adjusting the sample doping ratio, employing electrolyte (HCl, Na 2 SO 4 , NaOH), electrode reusability, applied temperature, and monochromatic illumination, the produced samples were used for efficient photoelectrochemical hydrogen production from water. Electrode stability, thermodynamic characteristics, conversion efficiencies, amount of hydrogen moles, and PEC impedance were also assessed and discussed. The 2.5% Co-doped SnO 2 photoanode that annealed at 673 K had the highest photocurrent (21.25 mA/cm 2 ), number of hydrogen moles (20.4 mmol/h.cm 2 ), IPCE (6.892%@307 nm), and APCE (4.61% at~500 nm). The optimised photoelectrode may be appropriate for industrial applications due to its excellent stability, high conversion efficiency, and inexpensive cost.