Design of SnO₂:Ni,Ir Nanoparticulate Photoelectrodes for Efficient Photoelectrochemical Water Splitting

Abstract

Currently, hydrogen generation via photocatalytic water splitting using semiconductors is regarded as a simple environmental solution to energy challenges. This paper discusses the effects of the doping of noble metals, Ir (3.0 at.%) and Ni (1.5–4.5 at.%), on the structure, morphology, optical properties, and photoelectrochemical performance of sol-gel-produced SnO₂ thin films. The incorporation of Ir and Ni influences the position of the peaks and the lattice characteristics of the tetragonal polycrystalline SnO₂ films. The films have a homogeneous, compact, and crack-free nanoparticulate morphology. As the doping level is increased, the grain size shrinks, and the films have a high proclivity for forming Sn–OH bonds. The optical bandgap of the un-doped film is 3.5 eV, which fluctuates depending on the doping elements and their ratios to 2.7 eV for the 3.0% Ni-doped SnO₂:Ir Photoelectrochemical (PEC) electrode. This electrode produces the highest photocurrent density (J_ph = 46.38 mA/cm²) and PEC hydrogen production rate (52.22 mmol h⁻¹cm⁻² at −1V), with an Incident-Photon-to-Current Efficiency (IPCE% )of 17.43% at 307 nm. The applied bias photon-to-current efficiency (ABPE) of this electrode is 1.038% at −0.839 V, with an offset of 0.391% at 0 V and 307 nm. These are the highest reported values for SnO₂-based PEC catalysts. The electrolyte type influences the J_ph values of photoelectrodes in the order J_ph(HCl) > J_ph(NaOH) > J_ph(Na₂SO₄). After 12 runs of reusability at −1 V, the optimized photoelectrode shows high stability and retains about 94.95% of its initial PEC performance, with a corrosion rate of 5.46 nm/year. This research provides a novel doping technique for the development of a highly active SnO₂-based photoelectrocatalyst for solar light-driven hydrogen fuel generation.

1. Introduction

Photoelectrochemical water splitting (PEC-WS) and the hydrogen generation process using unlimited solar energy have attracted increased attention worldwide. Hydrogen energy is extensively used in various fields, such as rocket propellant, vehicle fuel, and refining petroleum products. This is due to its zero CO₂ emissions, low density, high energy density, environmental friendliness, renewability, and storage capacity. This technique requires a large-area photoelectrode and a low-cost chemically stable material with significant absorption of solar light and a suitable bandgap [1,2,3,4].

Demonstrating abundance, facile preparation, chemical stability in a wide range of pH values [1], and environmental compatibility [5], SnO₂ is a semiconductor material that inherits a direct bandgap (E_g) of 3.6 eV with a high exciton binding energy at room temperature (RT) [2]; its electron mobility is in the order of 240 cm² V⁻¹ S⁻¹ [6], and it has high transparency in the visible range, low resistance, and high reflectivity for infrared light [7]. This oxide’s characteristics make it ideal for optoelectronics, solar cells, gas sensors, catalysis, and thin-film transistors (TFTs). Sn⁴⁺ (4d¹⁰5s⁰) is suggested to replace In³⁺ in thin-film transistors to avoid the high cost of device fabrication [8]. On the other hand, and despite these many advantages of SnO₂, its wide E_g and low surface area limit its use in industrial applications such as hydrogen generation.

Various physical and chemical approaches have been reported for the fabrication of doped and pure SnO₂ films and nanopowders. Sb-doped SnO₂ films were synthesized by means of the dip-coating technique [7] and spray pyrolysis [9] for possible use in optoelectronic applications and photocurrent generation. A SnO₂/ graphene sheet prepared by means of a hydrothermal method showed photocatalytic activity in dye removal, and this composite was found to be suitable as a supercapacitor electrode [10]. Bera et al. [11] synthesized uniform SnO₂ films of nanograss structure and SnO₂/BiVO₄ core-shell structures by means of a vapor-solid deposition process to achieve high photoelectrochemical efficiency. Moreover, Mohammad et al. [5] fabricated SnO₂-g-C₃N₄ nanostructures by means of the hydrothermal route for dye removal and wastewater treatment. Yuan et al. [12] found that WO₃@SnO₂ core-shell nanosheets with a 20 nm SnO₂ shell layer prepared using atomic layer deposition have superior sensing performance with a response of 1.55 and selectivity toward 15 ppm NH₃ at 473 K.

Among the used elements, the transition metals represent a fascinating class of materials as dopants. Doping with Co was reported to enhance the hydrogen gas sensing properties of the spin-coated SnO₂ films [3]. Cr doping was applied to tune the linear and nonlinear optical parameters of the spray-deposited SnO₂ thin films [13]. Moreover, doping SnO₂ films with Fe (5–20%) caused the transition of the conductivity of SnO₂ from n- to p-type and induced a good rectifying character for SnO₂ homojunction [14]. Ni doping has been reported to induce a greater number of defects and oxygen vacancies, resulting in an RT ferromagnetic character for the co-precipitated SnO₂ nanoparticles (NPs) [15]. Additionally, the Ni-doped SnO₂ was extensively studied for gas sensing applications [16,17]. Additionally, co-doping or double doping seems to be required for tuning the materials’ properties. Saadeddin et al. [18] found that the double doping of SnO₂ by Zn and Sb significantly improved its electrical conductivity and density, and the presence of Zn prevents the Sb evaporation during solid-state reaction at 1300 °C. According to Lamrani [19], the double (Eu, Gd) impurities substitute the adjacent Sn sites and result in strong ferromagnetic interactions via the p–f hybridization between rare-earth 4f and Op states. The obtained ferromagnetic alloy is expected to act as a great power source for effective photovoltaic conversion in solar cell applications. Moreover, Sery et al. [20] proved that co-doping of SnO₂ NPs with (3% Ti, 3% Ni) and co-doping of NiO NPs with (3% Ti, 3% Sn) significantly reduced their E_g and the obtained oxides are highly efficient photocatalysts.

The present work was focused on evaluating the influence of two transition metals, the noble metals Ir and Ni, on the physicochemical features of SnO₂ material. This work represents the first attempt to improve the structural, optical, and photoelectrochemical characterization of SnO₂:Ni,Ir films to enhance their potential for hydrogen energy generation. The samples were prepared by means of a sol-gel method, owing to its low cost, simplicity, safe, mass/large-scale production, and easy tailoring of film thickness. Additionally, sol-gel-prepared films usually have a mesoporous structure or a high specific area, which made these films more suitable for gas sensing and hydrogen energy production [3].

2. Experimental Details

2.1. Chemicals and Preparation Process

The source of Sn was SnCl₂·2H₂O, MW~225.63, from Merck, whereas the source of Ir element was IrCl₃·H₂O, MW~298.58 g/mol from Sigma. The Ni source was Ni(CH₃COO)₂. 4H₂O, MW ~ 248.84 from Fairsky. High-purity and water-free ethanol was used as a common solvent, and glacial acetic acid was used as a stabilizer. Meanwhile, 0.325 M solutions of SnO₂ and 3.0% Ir-doped SnO₂ (SnO₂:Ir) were synthesized by dissolving the required amounts of SnCl₂·2H₂O and IrCl₃·H₂O/SnCl₂·2H₂O, respectively, in 10 mL ethanol for each. These two solutions were then stirred at 60 °C for 3 h. After the first 20 min, a predetermined amount of the stabilizing agent was added drop by drop to obtain a clear and homogeneous solution. Then, the solutions were aged for one day at RT. To obtain SnO₂:Ni,Ir films, Ni(CH₃COO)₂·4H₂O was added to the second solution with specific weighted amounts for 1.5%, 3.0%, and 4.5% Ni-doped SnO₂:Ir structures. The spin coating process was carried out on pre-cleaned and dried glass substrates, and at 2500 rpm for 25 s, followed by drying for 15 min at 200 °C. The coating and drying processes were repeated 7 times. Finally, the films were subjected to heat treatments at 450 °C for 3.0 h in a controlled air furnace followed by cooling the furnace to RT overnight.

2.2. Characterization and Measurements

The crystal structure identification of doped and pure SnO₂ was evaluated using a high-resolution X-ray diffractometer PANalytical X’Pert Pro system, Holland. This was done using CuKα radiation of wavelength λ = 1.5406 Å with a step of 0.02° and operated at 42 kV and 32 mA. The patterns were recorded in the range of 4–70° on the 2θ-scale. The films’ surface morphology, thickness, and root mean square roughness (R_rms) were determined by means of atomic force microscope (AFM), the PARK SYSTEM, and XE-100E, respectively. Fourier transform infrared spectroscopy (FTIR) was used to investigate the films’ functional groups, utilizing FT/IR-6200 (Jasco Co., Tokyo, Japan), in a wavelength range 400–4000 cm⁻¹. The absorbance spectra (Abs) were recorded in a wavelength range of 200–1000 nm using a UV/VIS/NIR 3700 (Shimadzu, Columbia, MD 21046 USA) double beam Shimadzu spectrophotometer. The absorption coefficient $\alpha =2.303 Abs./film thickness \left(d\right)$ was used in calculating the direct optical E_g of the samples. All these investigations were performed at RT. The photoelectrochemical water splitting measurements were performed using 100 mL of 0.5 M HCl, NaOH, or Na₂SO₄ solutions at RT (25 °C) with a sweep rate of 1 mV/s. Only the temperature effect was investigated using a 0.1 M HCl solution. In the water-splitting test reaction, a two-electrode cell was used, with the photocathode being a nanocomposite electrode with a 1 cm² surface area and the counter electrode being a Pt electrode. The artificial light was incident on the electrode surface with standard white light illumination of 100 mW·cm⁻² provided by a 500 W Xenon lamp (Newport, 66142-500HX-R07, Newport, UK). The effects of incident wavelength (307–636 nm), temperature reaction (25–85 °C), and time stability on water splitting were investigated.

3. Results and Discussion

3.1. X-ray Diffraction and AFM Analysis

Figure 1a shows X-ray diffractometer (XRD) charts of doped and pure films. The patterns are composed of well-defined peaks with orientations in the (1 1 0), (1 0 1), (2 0 0), (2 1 1), (2 2 0), and (0 0 2) planes that match the standard data for a rutile phase of cassiterite SnO₂ with a tetragonal structure, JCPDS no. 72-1147, and space group P4₂/mnm. Among them, the (1 1 0) peak exhibits the highest intensity, and this direction has the least surface energy, and so is thermodynamically and electrostatically the most stable [21]. Moreover, the crystal growth in the (2 0 0) and (2 1 1) directions means that all films are polycrystalline and composed of randomly oriented crystals. No IrO_x or NiO phases are observed in the spectra under the resolution limit of XRD. A similar result was reported for hydrothermally prepared graphene-loaded SnO₂ NPs [10]. Figure 1b illustrates that the positions of (1 0 1) and (2 0 0) are influenced by doping, where the peaks are right-shifted with an increasing Ni doping ratio. These peaks’ shift confirms that the Ir/Ni ions were introduced substitutionally and/or interstitially inside the SnO₂ lattice [22]. Right-shifted diffraction peaks were also noticed in the XRD of spin-coated SnO₂ after loading with Co [3]. Additionally, the introduction of Zn²⁺ (0.74 Å radius) and F ions (1.33 Å radius) caused lattice expansion for the spray-deposited SnO₂ films [23].

The crystallite size (D) and lattice parameters (a,c) were determined from Scherrer’s equation, $D=0.9\lambda /({\beta }_{1/2}\mathit{cos}\theta )$, and the following formula [22]:

$$\frac{1}{d_{hkl}}=\frac{2 sin\theta }{\lambda }=\sqrt{\frac{\left(h^2+k^2\right)}{a^2}+\frac{l^2}{c^2}}$$

(1)

where ${\beta }_{1/2}$ is the full width at half maximum intensity, d is the inter-planar distance, and (hkl) is Miller’s indices. The calculated values of a, c, and the volume of the unit cell (V = a²c), as well as the crystallite size (D_av), are tabulated in

Table 1. Films’ thicknesses, XRD data (lattice parameters a and c and volume of the unit cell V), and crystallite size D_av. AFM measurements (grain size, Gs, and roughness, R_rms) and the direct optical bandgap (Eg) of the films) are also shown.

Film’s Composition	Thickness (nm)	Lattice Parameter			Dav (nm)	AFM		Eg (eV)
		a (Å)	c (Å)	V (Å3)		Gs (nm)	Rrms (nm)
SnO2	222	4.747	3.189	71.86	29.62	33.12	17.25	3.50
SnO2:Ir	219	4.742	3.191	71.75	30.43	35.07	15.36	3.70
SnO2: 1.5% Ni,Ir	225	4.741	3.182	71.52	27.57	29.37	19.17	2.77
SnO2: 3.0% Ni,Ir	217	4.735	3.178	71.25	23.91	26.92	14.23	2.70
SnO2: 4.5% Ni,Ir	215	4.712	3.173	70.45	24.23	23.81	11.64	2.75

. As apparent from Figure 1, the peaks’ intensity reduced with an increasing dopant concentration, indicating the decrease in the crystallinity of the films. The value of D_av was calculated from the most three intense peaks; (1 1 0), (1 0 1), and (2 1 1). D_av slightly increased from 29.62 to 30.43 nm after 3.0% Ir doping but decreased to 24.23 nm after 4.5% Ni loading. This means that Ir doping improved the film’s crystallinity. The decrease in D_av of SnO₂ with increasing dopant contents may be attributed to the formation of a large number of nucleation centers causing the formation of smaller nanocrystals [24]. The greater the density of centers of nucleation, the smaller D_av will be, and vice versa [25]. Similarly, D_av decreased from 28.3 nm for pure SnO₂ nanorods to 6.2 nm after 5.0% Ni doping [16]. As listed in Table 1, the calculated values of a, c, and V of pure SnO₂ are 4.747 Å, 3.189 Å, and 71.86 ų, respectively. These values are close to those reported by Bouznit and Henni [9], a = 4.75 Å and c = 3.18 Å for the spray-deposited SnO₂ film. Additionally, the parameters of the prepared samples are influenced by the dopant type and its content inside the SnO₂ film. When increasing the Ni content from 1.5% to 4.5%, the a, c and V are decreased from 4.741 Å, 3.182 Å, and 71.52 ų to 4.712 Å, 3.173 Å, and 70.45 ų, respectively. The decrease in these parameters is attributed to the replacement of the Sn⁴⁺ ion (radius ~0.72 Å) with smaller sized ions (Ir³⁺, 0.68 Å; Ni²⁺, 0.62 Å), as well as the associated interfacial effects at the dopant sites, such as charge redistribution and valence state regulation [26,27]. This also may be related to the nanoparticle size effect, whereas the lattice parameter contraction is proportional to the diameter of the nanoparticle. Singh et al. [2] reported a reduction in D_av and an increase in V of the co-precipitated SnO₂ NPs due to doping with Er at 1.0–5.0% ratios.

Figure 2 shows 3D and 2D AFM images for the doped and pure SnO₂ films. As noted, the substrates are well-covered with the material, and films are homogeneous, compact, and crack-free with a granular or particulate character. The surface morphology slightly changed after Ir loading, but depends on the Ni doping level. The white spots are nanocrystal islands with heights exceeding 6 nm in SnO₂ film and 10 nm in the Ni-doped IrSnO₂ films. Notably, 3.0% Ni loading changes the surface of SnO₂:Ir to a 1D tubular nanostructure, which is attractive for solar conversion via offering direct pathways for the charge carrier transport [6]. The films’ thickness is in the range of 215–225 nm, as listed in Table 1, i.e., the average of this range is ~220 nm. The grain size (Gs) and root mean square roughness (R_rms) of all films are listed in Table 1. The Gs marginally increased after introducing Ir³⁺, due to its comparable radius with that of the host (Sn⁴⁺). Then, the Gs decreases from 35.07 nm for SnO₂:Ir film to 23.81 nm with an increasing Ni doping ratio from 0 to 4.5%. The difference in radii between host Sn⁴⁺ and dopant Ni²⁺ may account for the decrease in Gs caused by increasing the Ni doping level. This behavior is consistent with the data obtained from XRD analysis, peak shift, lattice parameters, and unit cell volume. Previously, it was mentioned in [17] that Ni is a grain growth inhibitor within the SnO₂ matrix. This result is consistent with XRD data. However, as AFM measures Gs as the apparent distance between two adjacent grains, and as one grain may consist of more than one crystallite, Gs is thus found to be greater than D_av. Moreover, the R_rms of SnO₂ decreased from 17.25 to 15.36 nm after doping with 3.0% Ir, then increased to 19.17 nm after co-doping with 1.5% Ni. Finally, it decreased to 11.64 nm with an increased Ni doping ratio. Similar results were reported for the sol-gel dip-coated (Fe, Ni)-doped SnO₂ films [28]. In that paper, Benkara et al. illustrated that the R_rms strongly depends on the preparation and dipping conditions as well as the different kinetics of Fe and Ni needed for the growth of SnO₂ nanoparticles. It is well-known that low R_rms reduces the carrier scattering inside the films, ensuring the high performance of TFTs [8]. Additional morphological investigation was carried out by taking cross-sectional images for the films, as shown in Figure S1 (Supplementary Materials). These images confirm the tubular structures of the prepared films after the insertion of Ni and Ir as dopant materials. Surface morphology, it is believed, plays an important role in improving the material’s PEC performance. As a result, Ni and Ir doping is expected to encourage the use of these films as photoanodes in electrochemical applications.

3.2. FTIR and UV-Vis Spectroscopy

In the absorbance mode, the FTIR spectra of SnO₂, SnO₂:Ir, and SnO₂:Ni,Ir samples are displayed in Figure 3. The absorption band at 445 cm⁻¹ is owing to the symmetric Sn–O–Sn vibration. The band at 604 cm⁻¹ is attributed to the antisymmetric Sn–O–Sn/O–Sn–O bridge groups of SnO₂ [26]. Doping with Ni/Ir results in a slight decrease for the first band and a slight increase for the second one. The SnO₂ lattice mode appears at 688 cm⁻¹, and this band is also influenced by foreign atoms [22]. The observed bands are similar to those reported by Kuantama et al. [29] for nanoporous F: SnO₂ prepared by the sol-gel combustion route. The very strong infra-red (IR) absorption that occurs at ~905 cm⁻¹ was assigned to Sn–OH vibration [30]. Besides the reduction in intensity with doping, this band is also slightly shifted to lower energies. Moreover, the IR spectrum of SnO₂ film displays a broad peak around 3410 cm⁻¹, indicating the presence of O–H on the film surface. The IR absorption spectrum is size- and morphology-dependent [28]. The film doped with 3.0% Ni shows more intensive absorption bands compared with the films doped with 1.5 and 4.5% Ni. This is due to the small crystallites size at this doping ratio (23.91 nm) and the change to one-dimensional (ID) tubular morphology, as seen in Figure 2. The observed red-shift and decrease in peaks’ intensity due to the Ni/Ir doping are powerful pieces of evidence of the Ir and Ni introducing into the SnO₂ cassiterite lattice. The observed peaks related to the presence of OH indicate the films’ ability to be used in aqueous solutions for electrochemical applications.

UV-vis spectra of the films are shown in Figure 4a. As seen, although Ir incorporation reduced Abs, there is a significant increase in Abs after introducing Ni into the IrSnO₂ matrix. Ni-doped films show an absorption band around 255 nm, shifted to lower wavelengths for SnO₂ and SnO₂:Ir films. This band is due to the photoexcitation of electrons from the valence band (VB) to the conduction band (CB) [15]. Moreover, the absorption edge is shifted from 380 nm for the pure film to longer wavelengths after doping with Ni. Similarly, compared to the electro-spun SnO₂ nanotubes (NTs), the SnO₂/SnS₂ showed an absorption edge that red-shifted combined with an enhanced absorption intensity [31]. Since visible light represents a large portion of the solar spectrum, photoanodes with strong absorption for the visible light (i.e., Ni-doped films) should have higher PEC-WS performance.

Figure 4b shows (αhν)² vs. hν, the Tauc plots for determination of E_g, which are obtained by linear extrapolation to the abscissa axis. As given in Table 1, the E_g of pure SnO₂ is ~3.5 eV, and 3.0% Ir incorporation resulted in a blue-shift for E_g to 3.7 eV, whereas a red-shift to 2.70 eV is caused by Ni loading. This is consistent with the introduction of electronic states in the bulk and the oxygen defects associated with the doping. Although the E_g of the pure film is smaller than that reported in the literature, Mohammad et al. [5] reported an E_g of ~3.19 eV for the hydrothermally prepared SnO₂ film, and Li et al. [31] reported an E_g of ~3.37 eV for SnO₂ nanotubes fabricated by the electrospinning technique. The Burstein–Moss effect could explain the widening in E_g due to Ir doping. The narrowing in E_g with Ni doping may be related to the random distribution of Ni impurities and formation of the density of states tails at the edges of the bands, which result in an increased number of band-to-tail and tail-to-tail transitions [23]. In other words, this is due to the sp-d spin-exchange interactions between the localized d electrons of Ni ion substituting the cation and the band electrons. This result suggests that Ir/Ni co-doping could be used to tune the Eg of SnO₂ for specific applications.

In the literature, the E_g of SnO₂ reduced from 3.87 to 3.79 eV after doping with 10% Sb [7] and decreased from 3.6 to 3.28 eV after doping with Cr at ratios ≤ 5 wt% [13]. Moreover, according to Othmen et al. [14], the incorporation of Fe (5–20%) decreased the allowed and forbidden direct E_g of spin-coated SnO₂. Ni doping can alter the E_g of SnO₂ NPs from 4 to 3.85 eV [15]. Additionally, Ni doping at 5.0% reduced the E_g of SnO₂ nanorods, prepared by a hydrothermal method, from 3.85 to 3.65 eV [16]. According to Salameh et al. [23], the E_g of SnO₂ decreased from 3.97 to 3.53 eV after Zn loading but increased to 4.45 eV after co-doping with F. Han et al. were also able to reduce the energy gap of sol-gel-prepared SnO₂ films from 3.8 to 3.1 eV by using Ti alloying and the oxygen vacancies created [32]. In contrast to our results, Ren et al. [8] found that the E_g of SnO₂ was 3.87 eV and was not changed even after doping with 10–50% Ga. Additionally, Ni doping was found to increase the E_g of the spray-deposited SnO₂ films from 4.03 to 4.09 eV [17]. Guo et al. obtained SnO₂ electrode with an energy gap of 1.74 V after (Ni, S)-co-doping [33]. Additionally, Matussin et al. reduced the band gap energy from 3.33 eV for SnO₂ to 2.08 eV for 10at% (Co, Ni)- SnO₂ [34]. Moreover, Jasim et al. reported the reduction in the E_g of SnO₂ from 3.7 to 2.5 eV for SnO₂:Co: Cu–H nanoparticles [35]

3.3. Photoelectrochemical H₂ Generation

3.3.1. Influence of the Used Electrolyte

One of the significant parameters that affect the PEC performance of the photocatalyst is the PEC cell electrolyte. Any alteration in the conditions of the electrolytes will cause several differences in the photocatalytic interfaces. Electrolyte choices need to be consistent with the applied PEC catalyst. Electrolytes must not have electrochemical reactions with the used photocatalysts, and their optical absorption bands cannot overlay with that of photocatalysts. The structure of electrolytes has a major influence on the PEC H₂ generation rate, which affects the diffusion of ions and thus changes the conductivity of electrolytes. The H₂ generation rate and PEC catalytic efficiency depend on the pH values and the conductivity of the electrolytes [36]. Conductivity is an essential property for selecting electrolytes to allow the transport of charges in the PEC process. The addition of active cations/anions (such as Na+/SO₄²⁻ or OH⁻) with sufficient concentrations will reduce the electrolyte impedance and promotes band bending. Dissimilar electrolytes of the same concentration (0.5 M)—a strong acid (HCl, pH = 0.3), a strong base (NaOH, pH = 13.69), and neutral salt (Na₂SO₄)—were employed to identify the major variations in the PEC performances of the studied photocathodes. The PEC measurements of the photoelectrodes were taken in 100 mL of each electrolyte at 25 °C (RT) with increment 1 mV/s. During PEC measurements, a working 1 cm² electrode and a counter 1 cm² Pt electrode were used. The measurement of photocurrent density–voltage (J_ph–V) curves was performed in standard white light illumination of 100 mW.cm⁻² provided by a 500 W Xenon lamp, as shown in Figure 5a–c. The J_ph’s value rose as the used bias increased and, surprisingly, the current density increased following the order: J_ph(HCl) > J_ph(NaOH) > J_ph(Na₂SO₄). The maximum J_ph value is 46.38 mA/cm² in HCl, 18.54 mA/cm² in NaOH, and 14.95 mA/cm² in Na₂SO₄ solution. So, HCl is the best electrolyte for the PEC process utilizing the 3.0% Ni-doped SnO₂:Ir photoelectrode. Similar behavior was reported for F-doped and In-doped SnO₂ [37]. This could be due to a slightly low pH value (0.30) that enhances the PEC reaction in an acidic medium with a high density of H⁺ ions [38]. The obtained J_ph value for 3.0% Ni-doped SnO₂:Ir photoelectrode is higher than that obtained for 1.5% Ni-doped SnO₂:Ir, 4.5% Ni-doped SnO₂:Ir, SnO₂:Ir, and SnO₂ under exposure to light. So, the optimized photocathode is 3.0% Ni-doped SnO₂:Ir, which gives the maximum photocurrent density in the HCl electrolyte.

On the other hand, the observed anodic current especially in the HCl electrolyte (Figure 5d) is ascribed to the oxygen and chlorine evolution reactions (OER and CER) in the HCl electrolyte (Peak I and Peak II) for SnO₂ and SnO₂:Ir electrodes and only OER (Peak I) for 1.5% Ni-doped SnO₂:Ir and 3.0% Ni-doped SnO₂:Ir electrodes. The current density and voltage position of peaks I and II are shown in Table S1 (Supplementary Materials). For Na₂SO₄ (Figure 5b), the anodic current at a voltage > 0.5 V is ascribed to the OER only. It is noted that the use of NaOH (Figure 5c) decreased the OER efficiency of the SnO₂:3% Ni,Ir sample, which was found to occur because OH⁻ ions fight for adsorption sites and the kinetics of the reaction may be deaccelerated. Therefore, in the different electrolytes, the recorded anodic currents were predominantly due to oxygen evolution depending on the electrolyte, and not to SnO₂:Ni,Ir dissolution. Hence, in the various electrolytes, the observed anodic currents were mainly caused by oxygen evolution depending on the electrolyte, and not SnO₂:Ni,Ir dissolution [39]. So, under OER conditions, Ni and Ir co-doping could help to improve PEC stability, as we will discuss later. It was previously stated that the band bending at a high doping level increased with anodic current density, resulting in a monotonic current density increase at pH ≤ 1 due to the driven electron flow through the band bending and ionization of the deep donor levels of oxygen vacancies [40]. Electron tunneling is considered the principal mechanism of anodic current density at low doping percentages [41].

3.3.2. PEC Behavior of the 3.0% Ni-Doped SnO₂:Ir Photocathode

Figure 6a shows the PEC characteristics of the 3.0% Ni-doped SnO₂:Ir photoelectrode utilizing 0.5M HCl in the dark and under the standard white light of 0.1 W/cm² illuminations utilizing a Keithley 2400 source measure unit. As shown, the current density increased with an increasing voltage to reach −5.27 and −46.38 mA cm⁻² at −1V in the dark and when subjected to white light, respectively. The significant dark currents are caused by the transfer of charges supported by ionic currents that come from the HCl electrolyte. Under the visible light illumination, the sharp increase in J_ph with a negative potential indicates that this electrode acts as a photocathode. Remarkably, this PEC electrode displays a current density of ∼−432 μA/cm² at zero V and photocurrent onset at >−0.196 V. The high value of the photogenerated current density in light is a pre-indicator for the high-performance PEC water-splitting. This could be due to the extension of E_g to the visible light region (2.7 eV) and the robust absorption of Vis/IR by Ni and Ir co-doping, in addition to the observed 1D nanotubular morphology, which speeds up the redox reaction rate and thus promotes the PEC reaction in addition to enhancing the tunneling activity of the photogenerated carriers [42,43].

The analysis of the PEC (J_ph–V) curve was performed under visible light for 12 runs to investigate its ability to be reproduced and the reusability of 3.0% Ni-doped SnO₂:Ir photoelectrode. These results are shown in Figure 6b. The successive repeated measurements led to a decrement in the photocurrent density from −46.38 to 44.04 mA cm⁻² after 12 sequent runs at −1 V in 0.5M HCl. Therefore, after 12 runs of the ruse, this electrode conserves almost 94.95% of its initial PEC performance, which indicates high stability and reusability. The PEC stability of the 3.0% Ni-doped SnO₂:Ir photocathode was studied for a long time in 0.5M HCl under standard visible light illumination and −1V between the Pt counter electrode and this photocathode. The dependence of J on time is depicted in Figure 6c. As seen, the J_ph sharply decreased within the first 2 s to ~3.23 mA·cm⁻². This rapid fall in J_ph could be due to limited photocorrosion occurring between the redox electrolyte and this photoelectrode [43]. After that, a small decrease in J_ph is seen before reaching −1.28 mA cm⁻², as a constant value, for a time of >300 s, i.e., the current density decreases until a minimum is reached, illustrating the continuous accumulation of uncompensated ionic space charge at the two electrodes until beginning the injection of electronic charge. This shows that the 3.0% Ni-doped SnO₂:Ir is photochemically stable and has a long service life for H₂ generation through the PEC water splitting. Moreover, the photocurrent density was measured under pulsed light utilizing the chronoamperometry technique.

Figure 7a shows the variability in the photocurrent density versus the elapsed time during the on/off light switching for the 3.0% Ni-doped SnO₂:Ir electrode. Figure 7a shows the time required to reach a current density of 6.3 mA/cm² for 15 successive on/off cycles of light. This current/time behavior was measured at −0.53 V. Additionally, the variation of the detection time with the number of runs is provided in Figure 7b. For runs from 1 to 15, the time required to reach 6.3 mA/cm² increased from 117 to 192 s. The photocurrent density decreases in the first two cycles and then reaches a steady and quasi-reproducible state after several on/off cycles of light, with no over-shooting at the beginning or end of the on/off cycle, as shown in Figure S2 (Supplementary Materials). The duration of Figure S2 is 35 s, which represents the electrode’s holding time in the light. This suggests that the electron’s diffusion path is free from grain boundaries, which can introduce traps to impede electron movement and slow down the photocurrent generation [42]. The stability test of the 3.0% Ni-doped SnO₂:Ir electrode under successive on/off illumination cycles was carried out for 5 h to ensure long-term PEC stability, as shown in Figure S3 (Supplementary Materials).

Theoretically, the calculated number of generated hydrogen moles from the water splitting through the PEC process using Faraday’s law is represented by Equation (2) [44]:

$${\mathrm{H}}_2(\mathrm{moles})= {{\displaystyle \int }}_0^{\mathrm{t}}\frac{J_{ph}\mathrm{dt}}{\mathrm{F}}$$

(2)

where F = 9.65 $\times$ 10⁴ C/mol (the Faraday constant) and t is the time of generation. From the recorded J_ph–time data (Figure 6c), the number of produced H₂ moles as a function of the time is presented in Figure 6d. The calculated hydrogen output rate for the 3.0% Ni/IrSnO₂ electrode is 52.22 mmol h⁻¹cm⁻², while practically it was 35.03 mmol h⁻¹cm⁻². The Faraday efficiency is thus 67.1% for 3.0% Ni/IrSnO₂.

3.3.3. Effect of Temperature and Thermodynamic Paramters

Figure 8a illustrates the influence of heating in the RT–85 °C range on the PEC J_ph–voltage of the 3.0% Ni-doped SnO₂:Ir photoelectrode in 50 mL of 0.1M HCl electrolyte. One can see in this figure that J_ph is increased from −13.29 to −46.07 mA cm⁻² by increasing T to 85 °C. The increase in J_ph with the increase in T may be due to: (i) increasing T of the photo-induced carriers enhancing the number of electrons in the CB and holes in the VB, thus improving the rate of redox reactions and J_ph; (ii) increasing T leading to an increase in the charge carrier’s mobility and thus the carrier’s lifetime, following the relation μ= q${\tau }_n$ /m*, where μ indicates the carrier’s mobility, and m* is the carrier’s effective mass. This results in a decrement in the rate of recombination of the produced carriers, which increases the rate of hydrogen generation; and (iii) the improvement of the minority carrier diffusion length and hence the photocurrent density, which is directly proportional to the square root of the absolute T according to the relation J_ph α L_diff =$\sqrt{\mathsf{\mu } \frac{k_B\mathrm{T} }{q}{ \mathsf{\tau }}_{\mathrm{n} }}$ [45].

In addition, it is necessary to evaluate thermodynamic parameters such as activation energy (Ea), enthalpy (ΔH*), and entropy (Δs*). The relationship between the reciprocal of the absolute T (1/T) and J_ph (rate of reaction) for the 3.0% Ni/IrSnO₂ electrode is depicted in Figure 8b. The Ea value is calculated from the linear fitting slope of Figure 8b based on the Arrhenius equation [46]:

$$\mathrm{Ln} \left(J_{ph}\right)=-\frac{\mathrm{Ea}}{\mathrm{R}}[\frac{1}{\mathrm{T}}]$$

(3)

where R = 8.314 J.K⁻¹.mol⁻¹, the universal gas constant. From Figure 8b, slope = −Ea/R and the Ea value for the 3.0% Ni/IrSnO₂ electrode is 17.598 kJ.mol⁻¹. Additionally, the values of ΔH* and ΔS* for the H₂ generation reaction are calculated based on the Eyring equation. The Ln (J_ph/T) versus (1/T) is displayed in Figure 8c. This equation takes the form [46]:

$$\mathrm{Ln}\frac{J_{ph}}{\mathrm{T}}=-\frac{{\mathsf{\Delta }\mathrm{H}}^{*}}{\mathrm{R}}· \frac{1}{\mathrm{T}}+\mathrm{Ln}\left(\frac{k_B}{\mathrm{h}}\right)+\frac{{\mathsf{\Delta }\mathrm{S}}^{*}}{\mathrm{R}}$$

(4)

where k_B = 1.38$\times$10⁻²³ J.K⁻¹ (Boltzmann’s constant) and h = 6.626$\times$10⁻³⁴ J.s (Planck’s constant). From the slope of the linear fitting, the ΔH* value for 3.0% Ni/IrSnO₂ is equal to 20.31 kJ.mol⁻¹. Additionally, from the intercept, the ΔS* value is ~−107.91 Jmol⁻¹K⁻¹. For comparison, the thermodynamic parameters are also estimated for SnO₂ and SnO₂:Ir electrodes using Figure S4 and Figure S5 (Supplementary data). For the SnO₂ electrode, the Ea, ΔH*, and ΔS* values are ~14.308 kJ.mol⁻¹, 13.685 kJ.mol⁻¹, and −66.189 Jmol⁻¹K⁻¹, respectively. For the SnO₂:Ir electrode, these values are changed to 16.971 kJ.mol⁻¹, 23.83 kJ.mol⁻¹, and −637.68 Jmol⁻¹K⁻¹, respectively.

3.3.4. Effect of Monochromatic Light Illumination and Conversion Efficiencies

Figure 9a shows J_ph versus the applied voltage under the monochromatic light illumination for the 3.0% Ni-doped SnO₂:Ir photocathode in 0.5M HCl solution at RT. A set of bandpass filters with wavelengths from 307 to 636 ± 10 nm were used. From the inset of Figure 9a, the highest photocurrent occurred at 307 nm with J_ph = 38.25 mA/cm². The least current was created at 470 nm with J_ph = 30.41 mA cm⁻². This J_ph–wavelength dependence may be correlated with the absorption behavior of the 3.0% Ni/IrSnO₂ photocathode at each wavelength and supports the photoelectrocatalytic response of the optimized photoelectrode for the H₂ generation process. Generally, this behavior confirms the comprehensive response of the 3.0% Ni/IrSnO₂ photoelectrode and its capacity to absorb a large portion of the solar spectrum in UV/visible regions.

The enhanced solar absorption of the 3.0% Ni-doped SnO₂:Ir photoelectrode and its application for efficient H₂ production from H₂O splitting is further ensured by calculating the incident photon-to-current conversion efficiency (IPCE) at diverse wavelengths (λ). The IPCE% is calculated at a fixed applied voltage (−1 V) by Equation (5) [44]:

$$\mathrm{IPCE}\%=1240·\frac{J_{ph}}{\mathsf{\lambda }·\mathsf{\rho }}· 100$$

(5)

where $\mathsf{\rho }$ is the illuminating light power density of the xenon lamp (Newport, 66142-500HX-R07, Newport, UK) as a function of the monochromatic light wavelength—see Supplementary Data, Figure S6 and Table S2. The variation of IPCE% with λ is signified in Figure 9b. The maximum IPCE% for the 3.0% Ni-doped SnO₂:Ir photoelectrode is ~17.43% at 307 nm, which matches the maximum absorption of the optimized electrode.

The optical loss effects, including the incident photon transmission (Tr) or reflection (R), were ignored in the IPCE calculations. To measure the internal quantum efficiency, which is often called the absorbed photon-to-current conversion efficiency (APCE) to correct optical losses, APCE%, the number of PEC generated carriers which contribute to the generated photocurrent/absorbed photon is determined using Equation (6) [47]:

$$\mathrm{APCE}(\mathsf{\lambda })=\frac{\mathrm{IPCE}\left(\mathsf{\lambda }\right)}{\mathrm{A}\left(\mathsf{\lambda }\right)}=\frac{\mathrm{IPCE}\left(\mathsf{\lambda }\right)}{1-\mathrm{R}-T_r}$$

(6)

where A(λ) refers to the optical absorption. Figure 9c illustrates the variation of APCE% as a function of the incident λ. Two featured values are observed in this figure for APCE%: 18.45% at ~307 nm and 15.2% around 430 nm. These results agree well with the observation from Figure 4, whereas two absorption peaks and two optical bandgaps are observed for the 3.0% Ni/IrSnO₂ photoelectrode.

When applying a small external voltage to the PEC cell, the added electric energy to the system must be deducted to determine electrode photocatalytic efficiency. For this purpose, ABPE% (the applied bias photon-to-current efficiency) can be used. The ABPE% values for the designed photocathode are calculated by Equation (7) [48]:

$$\mathrm{ABPE}\%=J_{ph}\frac{\left(1.23-V_{app}\right)}{\mathsf{\rho }}\times 100,$$

(7)

where 1.23 is the standard state reversible potential of H₂O and V_app is the external voltage. Figure 9d illustrates the ABPE% values versus voltage at different wavelengths. Table S3 (Supplementary Materials) shows the maximum values of ABPE% and corresponding voltage (E) at different monochromatic light levels for the H₂ evolution reaction (HER) and the oxygen evolution reaction (OER). Figure S7 (Supplementary Materials) illustrates the ABPE% values as a function of applied voltage at different λ for OER. The maximum value of ABPE% is 1.038% at −0.839V and 307 nm for HER. Additionally, this photoelectrode shows an offset ABPE% of 0.391% at 307 nm and 0 V. This illustrates the decrease in the interfacial transport resistance and enhances the PEC performance. Moreover, two maximum ABPE% values are observed for OER (Table S3 and Figure S7): 0.433% at 0.166 V and 0.430% at 0.297 V at 307 nm. Note that the potential position of OER1 and OER2 shifts from 0.0653 to 1.66 V and from 0.351 to 0.297 V by decreasing the wavelength to 307 nm (Table S3). The high performance of the photoelectrode at relatively low potential could be advantageous for PEC cell setup.

The benchmark efficiency, or solar to hydrogen efficiency (STH), is the ratio between the total H₂ energy generated and the total energy of sunlight (AM 1.5 G, 0.1 W/cm⁻²). STH% can be measured utilizing Equation (8) [49]:

$$\mathrm{STH}=[ ({\mathrm{H}}_2/\mathrm{S})\times (2.37 \times {10}^5 \mathrm{J}/\mathrm{mol})]/[ \mathrm{A}\times {\mathrm{P}}_{\mathrm{total}}]$$

(8)

where P_total and A are the power density of the illuminating light in mW/cm², and the area of the illuminated part of the photoelectrode in cm², respectively, and H₂/S is the rate of the hydrogen moles production in mmol/s. Then, the calculated STH value is 12.72% for the 3.0% Ni-doped SnO₂:Ir photocathode.

In this study, the optimized 3.0% Ni-doped SnO₂:Ir photocathode showed higher J_ph and conversion efficiency than previously reported SnO₂-based PEC electrodes [50,51,52,53,54,55,56,57,58], as given in

Table 2. Values of J_ph and conversion efficiency of the optimized 3.0% Ni/IrSnO₂ photoelectrode relative to the previously reported SnO₂-based PEC electrodes for H₂O splitting.

Photoelectrode Photoelectrode	Electrolyte	PEC Performance	Ref
3.0% Ni-doped SnO2:Ir	0.5M HCl	Jph = 46.38 mA·cm−2 at −1V, IPCE = 17.43% and APCE% = 18.46% at 307 nm, 52.22 mmol h−1 cm−2 at −1 V	This work
SnO2-TiO2 heterostructure	0.25 M Na2S and 0.35 M Na2SO3 (pH ∼13)	Jph = 4.7 mA/cm2	[50]
Graphene/ SnO2–TiO2 heterostructure		Jph = 5.6 mA cm−2
Pt/SnO2/RuO2	0.1 M H2SO4	Rate of H2 production = 3.75 × 10−5 mol dm−3	[51]
FTO/SnO2 /W:BiVO4	0.15 M K2SO4 (pH 7)	1.4 mA/cm²	[52]
FTO/SnO2/W:BiVO4 /Co-Pi	0.1 M KPi (pH 7)	2.3 mA/cm²	[53,54]
Fluorine-Doped SnO2 Inverse Opal(IO)	0.5 M Na2SO4	Jph = 0.14 mA/cm2 at 1.23 VRHE, IPCE = 3.42% at 315 nm	[55]
F-Doped SnO2 /CuWO4 (4 cycles) IO film		Jph = 0.42 mA/cm2 at 1.23 VRHE, IPCE = 9.4% at 315 nm and 7.45% at 390 nm
BiVO4/WO3/SnO2 triple layers	0.5 M phosphate buffer electrolyte (pH 7)	Jph = 3.1 mA/cm2 at 1.23 VRHE, STH = 3.5%	[56]
FTO/BiVO4 /SnO2 /WO3 (double stacked)	0.1 M KHCO3 (pH 8)	Jph = 4.15 mA/cm2 at 1.23 VRHE	[57]
Ti/IrO2/Sb-SnO2 electrode	0.5 M Na2SO4	Jph = 0.4 mA/cm2 at 3 V	[58]

. Previous reports showed that the PEC J_ph of a graphene/SnO₂–TiO₂ heterostructure and FTO/BiVO₄/SnO₂/WO₃ was J_ph = 5.6 mA·cm^–2 at 1.23VRHE in 0.25 M Na₂S and 0.35 M Na₂SO₃ and J_ph = 4.15 mA·cm^–2 at 1.23VRHE in 0.1 M KHCO₃, respectively [50,57]. For a F-doped SnO₂/CuWO₄ (4 cycles) IO photocatalyst, the IPCE = 9.4% at 315 nm and 7.45% at 390 nm in 0.5 M Na₂SO₄ [55]. Additionally, an STH of 3.5% was reported for BiVO₄/WO₃/SnO₂ triple layers [56]. The results demonstrate that the nanostructured 3.0% Ni-doped SnO₂:Ir has high light-harvesting capabilities, high conversion efficiencies, and enhanced nanomorphological and nanostructural properties, which are highly favorable for practical applications.

3.4. Corrosion, Tafel Parameters, and Electrochemical Surface Area

The combined anodic and cathodic Tafel curves are presented based on the Tafel relation, V = β log(J_ph) + c, to outline the HER mechanism and the rate-limiting phase [59]. Perfect PEC catalysts have low Tafel slopes, high current exchange rates, and therefore, good HER efficiency. However, this is not a constant rule, where some photocatalysts with higher Tafel slopes and higher current conversion rates, and vice versa, exist [60]. Figure 10a illustrates combined anodic and cathodic Tafel plots of the electrodes under analysis (potentiodynamic polarization curves). Figure 10b presents the key characteristic parameters for the electrodes; corrosion potential (E_corr), corrosion current (I_corr), anodic and cathodic Tafel slopes (βa and βc). For the electrodes under analysis, the βa and βc values are determined from the slopes of the linear segments of the curves, as seen in Figure 10c,d [46,61].

For all electrodes, the determined values of E_corr, I_corr, βa, and βc are listed in

Table 3. Values of corrosion and Tafel parameters, and ECSA of our photoelectrodes.

Sample	ECorr (mV)	ICorr (μA cm−2)	βa (mV dec−1)	R2	Βc (mV dec−1)	R2	Rp (Ohm/cm−2)	Corr Rate (nm year−1)	ECSA (m2 g−1)
SnO2	−296	480.84	102.1 ± 4.4	0.9999	93.16 ± 0.23	0.9723	58.90	6.18	44.69
SnO2:Ir	−311	527.23	125.2 ± 5.3	0.9993	95.86 ± 0.64	0.9851	72.94	5.56	36.90
1.5% Ni doping	−282	163.29	90.4 ± 2.4	0.9983	50.35 ± 0.52	0.9943	142.82	1.69	16.02
3.0% Ni doping	−265	394.91	103.0 ± 6.1	0.9987	70.26 ± 1.03	0.9682	57.11	5.46	65.10
4.5% Ni doping	−306	517.13	105.8 ± 8.4	0.9995	104.09 ± 0.57	0.9530	73.31	5.35	41.34

. The βa and βc values are raised by adding 3% Ir to reach 125.2 and 95.9 mV/dec and then decreased by co-doping with 1.5% Ni to 90.4 and 50.4 mV/dec, respectively, whereas their values are increased to 105.8 and 104.1 mV/dec, respectively, by increasing the Ni content to 4.5%. Tafel PEC slopes clarify what the reactions mechanisms and rate-limiting stages are in the PEC process. That is, the Volmer–Tafel mechanism is dominant when the recombination stage is rate-limiting, and the Tafel slope is ~30 mV/decade. The Volmer–Heyrovsky hydrogen evolution mechanism can be presumed dominant when PEC desorption is rate-limiting, and the Tafel slope is ~40 mV/decade. If the slope of the Tafel is ~120 mV/decade, the reaction paths depend on the surface imbued with adsorbed hydrogen. The βc value demonstrates the over-potential needed to boost the hydrogen generation reaction (HGR) rate by a factor of 10 [46,61]. For the photoelectrodes under study, low Tafel slopes of the co-doped photoelectrodes and the low energy of the bandgaps resulted in low overpotentials. This is owing to the small amount of energy needed to improve the HGR performance and vice versa. Hence, the best values of βc and βa are obtained up to 3% co-doping. This explains the high PEC performance of our photoelectrode compared to the other photoelectrodes.

The information about the solution’s corrosion tendency can be obtained from E_corr, whereas the corrosion rate is directly proportional to I_corr. It can be observed from Table 3 that the 3.0% Ni-doped SnO₂:Ir photoelectrode showed better behavior with an E_corr of −265 mV and an electrochemical surface area (ECSA) of 65.10 m²/g. The 3.0% Ni-doped SnO₂:Ir has a higher E_corr (311 mV) in comparison with the 1.5% Ni-doped SnO₂:Ir and 3.0% Ni-doped SnO₂:Ir. Relative to the pure SnO₂, the corrosion potential of 3.0% Ni-doped SnO₂:Ir is moved towards the positive direction by 31 mV. Generally, the reported E_corr values are shifted to better behavior compared to that of the commercial SnO₂. Moreover, they are more positive than any reported data for photoelectrodes based on SnO₂ [62,63].

The values of I_corr, corrosion rate (CR), and polarization resistance (RP) can be used to determine the photoelectrode’s relative capability to resist corrosion. The CR is directly proportional to the kinetic value of I_corr, and Rp is inversely proportional to the kinetic value of I_corr. As seen from Table 3, the incorporation of 3% Ir decreases the I_corr of the pure SnO₂ from 359.6 to 323.6 μA/cm². Additionally, the insertion of 1.5% Ni decreases the I_corr of SnO₂:Ir to 98.5 μA·cm⁻². The increment of Ni content, however, increased the I_corr to 318.0 μA·cm⁻² at 3% Ni, but this is still much smaller than the corrosion current of pure photoelectrodes. The polarization resistance (Rp in Ohm/cm⁻²) was estimated from the relation; Rp = βc βa / [2.303 I_corr (βc + βa)] utilizing the Stern–Geary equation and the straight segment of the curves, close to E_corr. CR was also determined (in nm/year) using CR = 3272 I_corr[EW/d.A], where d, A and EW are the density (g/cm³), area (cm²), and the equivalent weight (g/eq), respectively. For all of our electrodes, the determined values of CR and Rp are given in Table 3. The value of Rp increases from 58.90 to 142.82 Ω.cm² and CR is reduced from 6.18 to 1.69 nm/year by incorporating 1.5% Ni/3% Ir to the pure SnO₂ photoelectrode. By increasing the co-doping ratio to 4.5%, the CR is raised marginally to 5.35 nm/year. The values of the corrosion parameters mentioned indicate the considerable enhancement of the corrosion resistance through the implementation of optimized Ni/Ir co-doping into the photoelectrodes. This further demonstrates the essential role of the Ni/Ir co-doping in improving the electrode’s stability. Our obtained values of CR are better than any previously reported SnO₂-based photoelectrode values [64,65]. Therefore, the corrosion parameters in Table 3 highlight the corrosion inhibition effects of 1.5% Ni-doped SnO₂:Ir, and the Tafel slopes of this photoelectrode also demonstrate a decrease in system dissolution reactions. This can be attributed to the more compact structure at this co-doping ratio due to the decrease in grain size and crystallite size and the development of a high density of smaller nanocrystals (XRD), in addition to the decrease in R_rms, which decreases the scattering of the carrier inside the electrode and electrochemical surface area and thus hinders corrosion.. This can thus be attributed to a more compact structure, a lower porosity, and a smoother surface. However, the increase in Ni to 3% changes the surface morphology of SnO₂:Ir from nanocrystals islands with heights ≤ 10 nm to a 1D tubular nanostructure, which is attractive for solar conversion by offering direct pathways for the charge carrier transport, the highest electrochemical surface area (65.1 m²/g), and the lowest optical band gap (2.70 eV) due to the insertion of electronic states in the bulk SnO₂ and co-doping-related oxygen defects.

The electrochemical surface areas (ECSAs) of the prepared photocatalytic electrodes are obtained using the Randles–Sevcik relation, ECSA = I(Rg T)1/2 (Ca N Fc)−3/2(v Da)-1/2/0.4463, wherever Rg, Ca, Fc, Da, and N indicate to the gas-molar constant, analyte concentration, and number of the electrons in the redox reaction (N = 1 in this study), respectively [46]. Utilizing Figure 5a, the ECSAs of the electrodes under analysis are calculated from ECSA = QH. mc⁻¹. Cm⁻¹, where QH, Cm, and mc are the negative-scan-hydrogen-adsorption charges after double-layered charge adjustment, the full charge of electrode-covering monolayer H-atoms, and catalyst mass, respectively [46]. The values of QH are determined from the integration of each photoelectrode’s curve divided by the measurement-scanning rate (10 mV) in Figure 5a. The obtained ECSAs are illustrated in Table 3. For the pure and 3.0% Ni/IrSnO₂, the ECSA values are 44.69 and 65.1 m² g⁻¹, respectively. The high ECSA indicate the high PEC performance of the 3.0% Ni-doped SnO₂:Ir photoelectrode compared to the other photoelectrodes [61].

3.5. PEC Impedance Spectroscopy (PEC-IS)

The impedance of the photoelectrochemical system depends on the charge transfer between the active electrode and the electrolyte interface. PEC-IS measurements were carried out at RT utilizing an electrochemical workstation (CH Instruments CHI660E) to examine the charge carrier dynamics of the optimized 3.0% Ni-doped SnO₂:Ir photoelectrode. The PEC-IS measurements were performed in a frequency (f) range of 10⁻²–10⁵ Hz at 0 V (vs. Ag/AgCl) under illumination. Figure 11a shows a Nyquist plot of 3.0% Ni/IrSnO₂ immersed in 0.5M HCl. The obtained results are also presented in Bode plots (Figure 11b,c). The Nyquist plot of 3.0% Ni-doped SnO₂:Ir photoelectrode shows that there is a small semicircle with a small diameter in the high-f side, indicating the redox reaction [66], followed by a linear stage covering the region of middle/low fs. These PEC-IS spectra displayed mixed diffusion/kinetic controlled routes. To understand the PEC-IS measurements through HER (Figure 11), the obtained spectra are fitted to a simple Randle equivalent circuit (REC). Figure 11a’s inset displays the REC that was used to model the PEC-IS spectra utilizing the ZSimpWin software. This equivalent circuit consists of electrolyte resistance (Rs = 2.76 Ω) obtainable at high fs from Nyquist plot intercepts, charge transfer resistance (R_ct = 3.87 Ω) equal to the semicircle diameter in the Nyquist plot, double-layer capacitance (C_dl = 19.63 mF) and Warburg impedance (W = 0.9). Then, the charge transfer process (CTP) is the main regulator of the HER, which is guaranteed by the uni-loop of the Nyquist plot. Although R_ct shows a higher value relative to the other elements of the equivalent circuit, this R_ct value is much smaller than most of the literature values for SnO₂-based PEC electrodes [55,67]. For example, F-doped SnO₂ IO and F-doped SnO₂/CuWO₄ IO showed significantly higher R_ct values of 18890 Ω and 8000–10000 Ω, respectively [55]. To the best of our knowledge, only the Ti/IrO₂/Sb-SnO₂ electrode showed an R_ct value of the same order (1.65 Ω), but with much smaller J_ph than our 3.0% Ni-doped SnO₂:Ir electrode [58]. The small value of R_ct, which represents the kinetic or charge transfer resistance to the ion transfer and is equal to the diameter of the depressed capacitive semicircle in the inset of Figure 11a, refers to enhanced ionic conduction and electrolyte diffusion over the electrode nanoporous structure [68]. Then, the optimized electrode can produce a large amount of H₂. Besides the charge transfer process (CTP), the e/h recombination is mostly used to control the HER. The reordered value of R_ct is very small, which means that the charge recombination on the electrode/electrolyte interface was highly reduced, which supports the improved HET [69].

Figure 11b,c shows Bode plots for the photocathode at the H₂ evolution potential in 0.5 M HCl aqueous liquid at 25 °C. Figure 11b shows the variation of the total impedance with f, and Figure 11c illustrates the variation of phase with f. This figure shows resistive regimes at low and very high frequencies, in addition to capacitive contribution in between. The low-frequency regime is correlated with the charge transfer resistance (R_ct) and the double-layered capacitance (C_dl) of the electrode. The very high-f regime may be related to the formation of a partially protective layer on the surface of the electrode. The maximum phase shift (θ_max= 16.4°) is detected at 14.68 Hz. The lifetime of the charge carriers can be estimated from Figure 11c by the relation ${\tau }_n$ = 1/2π ƒmax [70]. The calculated lifetime for the optimized electrode is 10.85 ms. This value is almost 31 times the reported value for pure SnO₂ (0.35 ms) [71]. Then, the determined parameters from Figure 11 indicate a great reduction in the charge recombination at the electrolyte/photoelectrode interfaces. This also refers to a kinetically facile PEC system, improved ionic conductivity, and electrolyte diffusion through the 3.0% Ni-doped SnO₂:Ir photoelectrode.

3.6. Sample Purity after Photoelectrochemical Measurements

XRD was measured for the optimized electrode, 3.0% Ni-doped SnO₂:Ir, after PEC measurements and is presented in Figure 12 relative to its XRD chart before PEC measurement. As seen, the (1 1 0), (101), and (2 0 0) peaks are slightly shifted to lower 2θ angles, indicating a marginal increase in d-spacing of the sample. The diffraction peaks became broader with reduced intensity, except for (2 0 0). The D_av decreased to 18.2 nm after PEC measurements. Only one small unknown peak is observed at 2θ = 44.03°. This result illustrates the high purity of the 3.0% Ni-doped SnO2:Ir photoelectrode.

4. Conclusions

SnO₂, SnO₂:Ir, and (1.5–4.5%) Ni-doped SnO₂:Ir films were deposited on a glass substrate using the spin-coating method. The results of XRD illustrate that all samples are polycrystalline structures with a rutile phase. The peak shift in XRD patterns and FTIR spectra confirmed the formation of homogenous composites. XRD and AFM showed that the films were compact, had granular structures, and that D (and GS) decreased from 30.43 nm (33.12 nm) to 24.23 nm (23.81 nm) with increasing Ni loading. Additionally, R_rms decreased from 17.25 to 11.64 nm. Ni-doped films show an absorption band at ~255 nm that shifted to lower wavelengths for pure and SnO₂:Ir films. Additionally, the direct E_g blue-shifted to 3.70 eV at 3.0% Ir doping, and then red-shifted to 2.7 eV for the Sn_0.94Ir_0.03Ni_0.03O₂ composition. The tendency of these films to form Sn–OH bonds, the decrease in R_rms, and E_g red-shifting to 2.7 eV are encouraging factors for H₂ generation from water. The key PEC factors for performance for pure, 3% Ir-doped, and Ni/Ir-co-doped electrodes were explored and improved for effective H₂ generation utilizing sunlight. Amongst the investigated PEC electrodes, the 3.0% Ni-doped SnO₂:Ir displays the highest photocurrent density of 46.38 mA/cm² and the highest PEC H₂ generation rate of 52.22 mmol h⁻¹cm⁻² at −1V with an IPCE% of ~17.43% at 307 nm. Additionally, this electrode showed an ABPE of 1.038% at −0.839 V and a surprisingly offset value of 0.391% at 0 V and 307 nm for HER, which are the highest values yet for SnO₂-based PEC catalysts. Moreover, this photoelectrode displays a photogenerated current density of −432 μA/cm² at 0 V and photocurrent onset over −0.196 V. The effect of the electrolyte type was studied, the current density was ranked as follows: J_ph(HCl) > J_ph(NaOH) > J_ph(Na₂SO₄). The βa and βc values are raised by adding 3% Ir to reach 125.2 and 95.9 mV/dec and then decreased by co-doping with 1.5% Ni to 90.4 and 50.4 mV/dec, respectively. As the T of the PEC process increases to 85 °C, the current density displays ~3.5-fold enhancement.

The thermodynamic factors have been obtained using the 3.0% Ni-doped SnO₂:Ir electrode; activation energy= 17.598 kJ/mol, entropy =107.91 J/mol^.K, and enthalpy = 20.31 kJ/mol. Additionally, this electrode demonstrated a higher electrochemical surface area (~ 1.8 times) and lower Tafel slopes (103.0 and 70.26 mv/dec) relative to the 3% Ir-doped photoelectrode. Moreover, the rate of corrosion changes from 6.18 to 1.69 nm year⁻¹ with the incorporation of 1.5% Ni/3% Ir co-dopants in the pure SnO₂ electrode. All of the key parameters demonstrated a significant decrease in charge recombination at the electrode/electrolyte interface, which was used to pinpoint the PEC H₂ generation mechanism.. From the stability study and after 12 runs of reusability at −1 V, the optimized photoelectrode preserved ~94.95% of its initial PEC performance. This work provided a new doping strategy to develop a new collection of highly active SnO₂-based photoelectrodes for sustainable PEC hydrogen fuel production under sunlight illumination.