Enhanced Photocatalytic Hydrogen Evolution from Water Splitting on Ta2O5/SrZrO3 Heterostructures Decorated with CuxO/RuO2 Cocatalysts

Photocatalytic H2 generation by water splitting is a promising alternative for producing renewable fuels. This work synthesized a new type of Ta2O5/SrZrO3 heterostructure with Ru and Cu (RuO2/CuxO/Ta2O5/SrZrO3) using solid-state chemistry methods to achieve a high H2 production of 5164 μmol g–1 h–1 under simulated solar light, 39 times higher than that produced using SrZrO3. The heterostructure performance is compared with other Ta2O5/SrZrO3 heterostructure compositions loaded with RuO2, CuxO, or Pt. CuxO is used to showcase the usage of less costly cocatalysts to produce H2. The photocatalytic activity toward H2 by the RuO2/CuxO/Ta2O5/SrZrO3 heterostructure remains the highest, followed by RuO2/Ta2O5/SrZrO3 > CuxO/Ta2O5/SrZrO3 > Pt/Ta2O5/SrZrO3 > Ta2O5/SrZrO3 > SrZrO3. Band gap tunability and high optical absorbance in the visible region are more prominent for the heterostructures containing cocatalysts (RuO2 or CuxO) and are even higher for the binary catalyst (RuO2/CuxO). The presence of the binary catalyst is observed to impact the charge carrier transport in Ta2O5/SrZrO3, improving the solar to hydrogen conversion efficiency. The results represent a valuable contribution to the design of SrZrO3-based heterostructures for photocatalytic H2 production by solar water splitting.


INTRODUCTION
ABO 3 is an inorganic perovskite with a mixed metal oxide composition, where the A-element is an alkaline (earth) or a lanthanide, and the B-element is a transition metal. An example of ABO 3 is zirconate (AZrO 3 ), known for its ferroelectric, piezoelectric, and photocatalytic properties. 1,2 In photocatalysis, the H 2 production efficiency of AZrO 3 remains low due to its limited visible light absorption (E g > 4 eV) and poor carrier generation. 3,4 Strategies to stimulate photocarrier generation as a means to improve H 2 water splitting under visible light are key for AZrO 3 . A way forward is producing a semiconductor via cation replacement (A = Ba, Ca, Sr) in AZrO 3 , followed by band alignment interfacing AZrO 3 with another semiconductor to form a heterostructure. 5−9 First, cation replacement can be done by introducing Sr into AZrO 3 to form SrZrO 3 , which has an orthorhombic crystal structure with a Pbnm space group. 10,11 SrZrO 3 is an indirect band gap semiconductor. The valence band (VB) lies lower than the water oxidation potential, while the conduction band (CB) is located higher than the hydrogen reduction potential. 12 Photogenerated carriers through VB and CB can recombine, reaching the SrZrO 3 surface and induce the chemical transformation of 2H 2 O into 2H 2 and O 2 . However, due to its wide band gap (E g ∼ 4 eV), 8 SrZrO 3 requires UV light to photogenerate enough carriers to produce 50 μmol g −1 h −1 . 6 The H 2 production can be improved to reach 5310 μmol g −1 h −1 using UV light and electron donor species, such as Na 2 S and Na 2 SO 3 . 7 Although the addition of electron donor species is an option, 7 the main challenge remains with the photocatalyst. An ideal catalyst should effectively promote charge transport and retain similar H 2 water-splitting performances under visible light.
The heterostructure concept involves band alignment, 13 which ideally can be used to modulate charge transport. This can be done by incorporating Ta compounds, such as Ta 2 O 5 and other tantalates, recognized as active photocatalysts for H 2 water splitting. 14 The band gap structure in tantalum oxide consists of O 2p orbitals formed by the VB and the CB, with a d 0 electronic configuration that provides electron mobility access. 15 Depending on the synthetic approach, 16 the addition of Ta can lead to doping via SrZrO 3 substitution or yield Ta segregates to form Ta 2 O 5 , especially when treated at high temperatures. 17 Notably, both Ta-substitution and Ta segregate formation can promote mobility access in photocatalysts. 18 However, H 2 water splitting in tantalates has been mainly promoted with UV irradiation. 15 From this aspect, the next desired step for Ta-containing SrZrO 3 catalysts is to retain charge transport properties under visible light. 5,6,8,19 This entails the increase of photocarrier density using visible light by coupling other chemical species, such as cocatalysts (or binary catalysts, hereafter bicatalysts), to Ta-containing SrZrO 3 . From this point of view, the heterostructure concept with the incorporation of a cocatalyst or bicatalyst has not been applied to Ta-containing SrZrO 3 yet, opening new opportunities to design SrZrO 3 -based photocatalysts. 20,21 Coupling a narrow band gap to a wide band gap semiconductor enhances light absorption in the visible spectrum. 22,23 In essence, this entails band gap tunability via band alignment to reduce the recombination of photogenerated charges. 24 Copper oxide can function as a narrow band gap p-type semiconductor (Cu 2 O), 25,26 catalyst (CuO), 26 or both, especially when Cu 2 O and CuO species are combined (hereafter, Cu x O). 27 It could then be expected to improve the exchange of photocarriers when interfaced with wide-band semiconductors enabling high catalytic activity. Furthermore, interfacing Cu x O with an oxide-based hydrogen evolution catalyst, such as RuO 2 , is an attractive option to improve H 2 production during water splitting. 23 The combination of Cu x O and RuO 2 has been successfully applied in photocathodes 23 and is now proposed to improve the photocatalytic activity of Ta heterostructure is also compared with Pt, a more costly catalyst than Ru or Cu. 28 In-depth chemical and structural analyses were carried out by X-ray photoelectron spectroscopy (XPS), electron energy-loss spectroscopy (EELS), and transmission electron microscopy (TEM) to understand the chemical states of the RuO 2 /Cu x O/Ta 2 O 5 / SrZrO 3 components. Ta 2 O 5 has been observed to be distributed at the surface and between grain boundaries in SrZrO 3 nanocrystallites, facilitating charge mobility during photocatalytic water splitting. Ru and Cu have been found as oxides, that is, RuO 2 , Cu 2 O, and CuO. RuO 2 has been seen to be shaped as nanorods over Ta 2 O 5 /SrZrO 3 , whereas Cu x O remains distributed over Ta 2 O 5 /SrZrO 3 with no particular shape. The photocatalytic activity of the heterostructure is attributed to a synergistic effect that allows charge transfer through energy channels, enabling charge carriers to recombine and reach the interface of the RuO 2 /Cu x O bicatalyst. To the best of our knowledge, this is the first report on the coupling of RuO 2 /CuO x to Ta 2 O 5 /SrZrO 3 for photocatalytic water splitting under visible light. Our results can contribute to the design of efficient SrZrO 3 -based photocatalysts for hydrogen evolution. 2.3. TEM, Energy-Dispersive X-ray Spectrometry, and EELS. Scanning transmission electron microscopy (STEM), energydispersive X-ray spectrometry (EDXS), and EELS were carried out using a Cs-corrected microscope JEOL ARM 200CF equipped with an JEOL SSD EDX spectrometer and a Gatan Dual EELS Quantum spectrum-imaging filter. The operational voltage was 200 kV. The photocatalyst powders were dispersed in ethanol and were deposited over different carbon-coated Au, Cu, and Ni grids before the inspection.

Chemical Analysis by XPS.
For the X-ray photoelectron spectroscopy (XPS) measurements, a Quantera SXM (Physical Electronics) was used. The X-rays were Al Kα, monochromatic at 1486.6 eV with a beam size of 200 μm. The binding energies were corrected according to the C 1s peak (284.8 eV). Samples were located on millimetric-sized indium cups, forming a pellet for sample homogeneity. In every sample, three different areas were probed with an area size of 600 × 300 μm 2 .
2.6. Optical Characterization. The optical properties were analyzed using a UV−vis NIR spectrophotometer (Cary 5000) in the diffuse reflectance mode. The band gap was calculated with the Tauc method, which involves plotting (α h ν) 1/n versus (h ν). The value of the exponent n denotes the nature of the sample transition, the value is 2, considering indirect allowed transitions. A linear region was used to extrapolate to the X-axis intercept to find the band gap values. Photoluminescence spectra were collected in an Agilent Cary Eclipse spectrophotometer using a 254 nm excitation. Prior to UV−vis-NiR or PL, the samples were sieved and pelletized.

Photoelectrochemical Characterization.
The photoelectrochemical measurements were carried out in a three-electrode quartz cell connected to a potentiostat from AUTOLAB. Pt was used as a counter electrode and Ag/AgCl (3 M KCl) as a reference electrode. The working electrode was fabricated by depositing the photocatalyst over an ITO substrate. For this process, 2 mg/mL of the photocatalyst suspension in ethanol was deposited using a spin coater at 2000 rpm. The samples were dried at 80°C for 10 min. Once dried, the samples are immersed in 0.5 M Na 2 SO 4 and used as an electrolyte. Electrochemical impedance spectroscopy (EIS) measurements for obtaining Mott Schottky plots were performed under dark conditions in a potential range of 0.8 to −0.8 V vs. Ag/AgCl at a frequency of 100 kHz−100 MHz and an AC perturbation of 10 mV.
The potential versus Ag/AgCl, E Ag/AgCl , was converted to reversible hydrogen electrode potential, E RHE , using the Nernst equation. For the photocurrent response experiment, a constant potential of 0.3 V vs. Ag/AgCl is applied. The electrode was illuminated with a solar simulator (Xe lamp 100 mW/cm 2 ) for 300 s, and the photocurrent was obtained considering the electrode area (1 cm 2 ).

Photocatalytic H 2 Evolution.
The photocatalytic experiments were performed in a Pyrex reactor of 250 mL. In a typical experiment, 0.1 g of the photocatalyst was dispersed in 200 mL of deionized water. Before each experiment, the reactor was purged with N 2 for 30 min and irradiated with a wide range UV−vis xenon lamp (simulated solar light). The photocatalyst was stimulated with irradiation between 400 and 900 nm at 100 mW/cm 2 in demineralized water. The oxygen and hydrogen products were analyzed using a gas chromatograph (Thermo Scientific) coupled with a thermal conductivity detector. No buffer or electrolyzer was added during the reaction, and the starting pH was 7. No external potential was applied during photocatalytic experiments.
The solar to hydrogen conversion efficiency (STH) was estimated from eq 1, 29 using the H 2 production, the Gibbs free energy for the reaction, the incident power of the solar simulator (100 mW/cm 2 AM1.5G), and the area of irradiation.
The quantum efficiency (QE) was calculated with eq 2, 29 at 420 nm, where N H2 is the number of H 2 molecules produced in seconds and N hv is the photon flux.  30,31 The revised Perdew−Burke−Ernzerhof for solids (PBEsol) were selected for cell-optimization as it reduces PBE's tendency to overestimate unit cell parameters. 32,33 The one-electron Kohn−Sham orbitals were expanded on a plane-wave basis with a kinetic energy cutoff for the plane waves of 800 eV (PBEsol calculations). PAW potentials were employed to describe the interaction between the valence electrons and the core electrons. 34 Reciprocal space integration over the Brillouin zone was approximated with finite sampling using Monkhorst−Pack k-point grids of 7 × 7 × 7. 35,36 The bulk unit cell of SrZrO 3 was optimized until the largest force on all atomic coordinates became smaller than 0.01 eV/Å. Furthermore, the convergence criterion for the self-consistent electric field (SCF) problem was set to 10 −6 eV for all optimizations, and the symmetry group was preserved throughout all simulations. The unit cell volume was kept fixed at different cell volumes, followed by a constant volume cell optimization to verify the strain effect on the band gap. The unit cell of both structures was scaled proportionally to investigate the effect of strain on the band gap. Furthermore, a band gap evaluation on the optimized PBEsol structures was performed employing the HSE06 37 hybrid functional and a kinetic energy cutoff of 550 eV using a k-point grid of 3 × 3 × 3 as well as similar electronic and force convergence criteria.

RESULTS AND DISCUSSION
A SrZrO 3 (Figure 1), followed by a discussion on the higher-order heterostructures, such as RuO 2 /Cu x O/Ta 2 O 5 /SrZrO 3 ( Figure  2). In Figure 1a, the morphology of SrZrO 3 consists of agglomerated particles of ten to hundreds of nanometer sizes with a uniform distribution of chemical elements Sr, Zr, and O. The crystal structure of SrZrO 3 is visualized along the [100] zone axis in Figure 1b,c that corresponds to the perovskite orthorhombic phase. An atomic model of the SrZrO 3 structure is depicted in Figure 1d. The identification and orientation of the crystal lattice planes are extracted from the fast Fourier transform (FFT) shown in Figure 1e.
In Figure 2a, the STEM-EDXS maps of 0.1%RuO 2 /1% Cu x O/3%Ta 2 O 5 /SrZrO 3 heterostructure show the distribution of Sr, Zr, and O, corresponding to the SrZrO 3 nanocrystallite formation. The Ru EDXS signal map indicated the growth of nanorods characterized in detail in Figures S1 and S2. The composition of the nanorods is RuO 2 (Figures S1 and S2), and they are distributed at various locations over the heterostructure, ranging in size from 10 to 30 nm in width and 100 to 200 nm in length. In the case of Cu, the overlapping signals of Cu Kα 8.04 with Ta Lα 8.140 (and Hf Lα 7.898 present as an impurity from the synthesis precursor) turned the EDXS mapping problematic for small quantities. However, when the amount of Cu is significant, it is possible to detect Cu among the SrZrO 3 nanocrystallites (see Figure S3). The Cu morphology is found not as distinctive as the RuO 2 nanorods but rather in the form of agglomerates, in a mixture state of CuO and Cu 2 O according to EELS observations ( Figure S3d).
The distribution of Ta is observed in various parts of the SrZrO 3 nanocrystallites: (i) dispersed over the SrZrO 3 nanocrystallites and (ii) accumulated in selected regions (Figures 2a and S4). A closer look at RuO 2 /Cu x O/Ta 2 O 5 / SrZrO 3 revealed that Ta segregated between the grains, as seen in the HAADF image in Figure 2b (see also Figure S4b). This can be distinguished by the higher contrast observed at the grain boundaries, corresponding to an accumulation of Ta (a higher Z = 73 element compared to Sr = 38 and Zr = 40). A similar observation in Figure 2c revealed Ta at the surface of the SrZrO 3 nanocrystallites (see also Figure S4d). To verify our hypothesis (and discard the presence of Hf Z = 72), EDXS and EELS are carried out in these distinct regions ( Figure S4). The Ta O 2,3 edge was detected when collecting the EELS signal from the high contrast region in the HAADF image ( Figure S4d); likewise, by performing EDXS in a similar area, the presence of the Ta Lα 8.140 peak was observed in the spectra (as shown Figure S4c). This detailed examination revealed that when Ta accumulates preferentially more in some grains than in others, it segregates at the grain boundaries and decorates the nanocrystallite surface. In addition, Ta is found forming clusters around the crystallites as seen in Figure 2d and confirmed by the Ta O 2,3 edge in the EELS signal in Figure  2e.  Figure 3 shows the XPS spectra of (a1−d1) Sr 3d, (a2− d2) Zr 3d, (a3−d3) Ta 4d, and (a4−d4) O 1s. The analyzed samples are displayed per row. In this case, (a1−a4) SrZrO 3 , (b1−b4) 3%Ta 2 O 5 /SrZrO 3 , (c1−c4) 1%Cu x O/3%Ta 2 O 5 / SrZrO 3 , and (d1−d4) 0.1%RuO 2 /1%Cu x O/3%Ta 2 O 5 / SrZrO 3 . Irrespective of the sample structure, the Sr 3d and Zr 3d core level XPS spectra show almost superimposable envelopes. The position of the Sr 3d 5/2 , Sr 3d 3/2 , Zr 3d 5/2 and Zr 3d 3/2 components located at 132.9, 134.7, 181.2, and 183.6 eV, respectively, indicate Sr 2+ and Zr 4+ in a SrZrO 3 environment (Table S1). 7,38 The specific area ratios (2/3) and spin−orbit splitting values for Sr 3d (1.8 eV) and Zr 3d (2.4 eV) suggest no secondary phase. For Ta 4d (a3−d3), unsurprisingly, the pure SrZrO 3 sample (a3) shows no Ta presence. The Ta 4d envelopes of the three other samples are identical and show two main contributions at 229.2 eV (Ta 4d 5/2 ) and 241.6 eV (Ta 4d 3/2 ) assigned to Ta 5+ in Ta 2 O 5 . 39−41 A less-resolved contribution is also observed at lower binding energies (ca. 224.3 eV) and is attributed to Ta 4d 5/2 of hydrated Ta species. Finally, the O 1s core-level XPS spectra (a4−d4) display broad envelopes that can be fitted with three components. The first contribution at lower binding energies, ca. 529.2 eV, is assigned to O 2− in metal oxides (i.e., SrZrO 3 , Cu x O, and RuO 2 ). The contribution at 531.2 eV is  attributed to oxygen adsorbed in SrZrO 3 , 7 while the contribution at the highest binding energies, ca. 532.6 eV, could be associated with O−H. 42 It can be concluded that there is no significant difference in the chemical environments of Sr, Zr, Ta, and O species for SrZrO 3 , 3%Ta 2 The Cu 2p and Ru 3p core-level XPS spectra of the heterostructure samples containing Cu and Ru, that is, 1% Cu x O/3%Ta 2 O 5 /SrZrO 3 (Figure 4a1,a2) and 0.1%RuO 2 /1% Cu x O/3%Ta 2 O 5 /SrZrO 3 (Figure 4b1,b2), are presented in Figure 4. Although the spectra have a low signal-to-noise ratio, the Cu 2p and Ru 3p peaks still provide valuable information. It should be noted that Ru 3d is not reported due to the elemental overlap with C, as observed in Figure S5. In Figures  4a1,b1, two contributions in the form of 932.7 and 934.2 eV peaks assigned to Cu 2 O and CuO are observed. 8,43 In this set of samples, the additional contribution at 942.4 eV is assigned to Cu 2p 3/2 satellites. 8,43 The coexistence of the Cu + and Cu 2+ oxidation states is corroborated by the Cu LMM spectrum ( Figure S5). The presence of the Cu 2 O and CuO phases is observed even after the water-splitting reaction ( Figure S5). The presence of Cu + and Cu 2+ also agrees with EELS measurement ( Figure S3). XPS confirms the presence of Ru in the 0.1%RuO 2 /1%Cu x O/3%Ta 2 O 5 /SrZrO 3 heterostructure (Figure 4b2). The binding energy of Ru 3p3/2 of ca. 463.0 eV agrees with the presence of Ru 4+ in RuO 2 (Table S1). 44 The chemical information, elemental composition, and chemical environments are summarized in Tables 1 and S1. The chemical environment of Sr and Zr and the Sr/Zr ratio are notably constant for all the studied heterostructures and unaltered even after the photocatalytic test ( Figure S5 and Table S1). However, a small reduction in Ta, Cu, and Ru is found after the photocatalytic water splitting for the 0.1% RuO 2 /1%Cu x O/3%Ta 2 O 5 /SrZrO 3 heterostructure (Table 1)  loadings. The inset in Figure 5a shows the Tauc plots estimated from the UV−vis spectra. Band gap for Ta 2 O 5 / SrZrO 3 has been found between 3.85 and 4 eV. A redshift to lower energies is observed for the highest Ta 2 O 5 -loaded samples. A reduction in the absorption band near a wavelength (λ) of 250 nm is seen in the UV−vis spectrum for 3 wt % Ta 2 O 5 , probably due to the participation of Ta 5d orbitals affecting the CB. 45 It should be mentioned that such an effect can promote charge separation, resulting in a significant benefit for a photocatalytic process. The results are in good agreement with photoluminescent (PL) measurements in Figure 5b, indicating a reduction in charge recombination for 3%Ta 2 O 5 / SrZrO 3 .
Light absorption in the visible range increases with Cu x O and RuO 2 in 3%Ta 2 O 5 /SrZrO 3 (Figure 5c). The results show a considerable increase in light absorption for the 0.1%RuO 2 /1% Cu x O/3%Ta 2 O 5 /SrZrO 3 heterostructure, which is more significant than those for 1%Cu x O/3%Ta 2 O 5 /SrZrO 3 and 1% RuO 2 /3%Ta 2 O 5 /SrZrO 3 . Therefore, it can be argued that the 0.1%RuO 2 /1%Cu x O/3%Ta 2 O 5 /SrZrO 3 heterostructure re-duces further charge recombination, as shown in Figure 5d. The results in Figure 5d suggest that by controlling RuO 2 / Cu x O ratios, visible light absorption can be optimized to maintain the photocatalytic rate high. 7 It should be noted that in Figure 5d, the PL spectrum of 3%Ta 2 O 5 /SrZrO 3 overlaps with the 1%Cu x O/3%Ta 2 O 5 /SrZrO 3 spectrum. Both spectra are also comparable to that of 1%RuO 2 /3%Ta 2 O 5 /SrZrO 3 .      structures is scaled proportionally to investigate the effect of strain on the band gap. Via subsequent constant volume optimization at PBEsol, 33,34 it is possible to verify the strain effect on the band gap. At the PBEsol level of theory, the band gap for both SrZrO 3 structures increases when applying compressive strain and decreases with tensile strain. Furthermore, a rigorous evaluation of the band gap using the hybrid functional of Heyd−Scuseria−Ernzerhof (HSE06) 37,46 is employed. It has been found that the HSE06 functional is superior in localizing valence electrons of transition metals (e.g., those in Cu 3d orbitals) more correctly than (semi)local density functionals. 47 An experimental band gap close to 5.6 eV for single SrZrO 3 crystals is typical, and HSE06 predicts theoretical band gaps of about ∼5.0 eV, 48 which is in line with the HSE06-calculated band gaps of 5.09 eV (Pnma) and 5.11 eV (Pnmb) in Figure 7. For all unit cell volumes, it is clear that the HSE06 calculated band gaps are higher than those obtained from PBEsol. However, the trend remains the same. The results suggest that strain effects may originate from the presence of Ta 2 O 5 after the synthesis procedure. Ta in SrZrO 3 induces compressive strain on the lattice, leading to lower unit cell volumes. Computationally, it has been found that compressive strain increases the band gap, while tensile strain leads to lower band gaps, as in low-loaded SrZrO 3 (Figure 5a). The effect is primarily due to (i) Zr +4 substitution by Ta +5 or (ii) strain effects on SrZrO 3 caused by segregated Ta 2 O 5 , both leading to a broader band gap.
In the case of 0.1%RuO 2 /1%Cu x O/3%Ta 2 O 5 /SrZrO 3 heterostructure, the presence of RuO 2 , Cu x O, or their combination RuO 2 /Cu x O leads to broader photoadsorption over a larger part of the visible spectrum (Figure 5c). The rationale behind this is that these oxides have lower band gaps compared to Ta 2 O 5 /SrZrO 3 (Figure 5a). 49 The measured UV−vis diffuse reflectance spectra (Figure 5c Figure 5c. Although there is a band gap difference of 0.4 eV between heterostructures with or without a bicatalyst, the role of Ta is imminent, either substituting Zr 4+ or compressive strain 19 in the SrZrO 3 lattice (Figure 7). It should be mentioned that no peak characteristics of RuO 2 , CuO, or CuO 2 have been found in the XRD pattern, possibly due to the low cocatalyst amounts used (lower than 5%).
In short, a detailed analysis of the heterostructure components and the effect of Ta 2 O 5 in SrZrO 3 has been carried out optically (Figure 5a). Ta 2 O 5 has a positive effect by lowering charge recombination, as indicated by the photoluminescent measurements in Figure 5b. The effect of Ta 2 O 5 in the SrZrO 3 structure leads to band gap tunability and has been studied further in Figures 6 and 7. The results show that the role of tantalum is imminent, by either substituting Zr 4+ or introducing compressive strain in the SrZrO 3 lattice. Lattice constraints in SrZrO 3 due to the presence of Ta are not observed in TEM, pointing toward shallow Ta 5+ doping. Although this lattice effect is not seen locally in Figure 2, the XRD pattern in Figure 6b reveals cell volume contraction for Ta 2 O 5 /SrZrO 3 . Therefore, Ta-substitution or ejected strain in SrZrO 3 should not be disregarded in Ta 50 For still larger Ta 2 O 5 loadings (i.e., 3.9 wt %), the H 2 evolution activity reduces to 959 μmol g −1 h −1 . The results indicate that Ta 2 O 5 loadings can also affect the overall catalyst performance. It can then be hypothesized that there is a trade-off between charge mobility 15 and trapped states for different Ta 2 O 5 loadings. From the results, Ta 2 O 5 in SrZrO 3 is maintained fixed to 3 wt %, as it shows the highest amount of H 2 produced in Figure 8a. It should be noted that during experiments shown in Figure 8a, the production of O 2 has not been observed.
The H 2 production is further improved by incorporating various RuO 2 /Cu x O loadings to 3%Ta 2 O 5 /SrZrO 3 . Insights on the kinetics of H 2 evolution on RuO 2 /Cu x O heterostructures are presented in Figure 8b. The results reveal that the H 2 production in the first 3 h shows a linear tendency. After this time, the production rate is diminished, showing a plateau effect, which several authors correlate to limitations in the surface area and the available active sites on the photocatalyst. 51 However, we should not disregard possible elemental losses after the reaction (Table 1). We also assess potential changes in the chemical environment and crystalline structure in 0.1%RuO 2 /1%Cu x O/3%Ta 2 O 5 /SrZrO 3 with XPS and XRD after the reaction (Figures S5 and S6 and Table S1). In this case, no significant changes are observed; only the Ta at % reduces nearly 2-fold at the surface (Table 1), possibly explaining the changes in Figure 8b after 3 h. The cumulative H 2 production is presented in Figure 8c. Figure 8c shows that 0.1%RuO 2 /1%Cu x O is the ideal ratio, yielding a H 2 production rate of 5164 μmol g −1 h −1 , which is even higher than those of several cocatalysts (e.g., RuO 2 , Cu x O, and Pt) and other zirconates and perovskite heterostructures as shown in Figure  8e and Table S2. The photocatalytic activity of 0.1%RuO 2 /1% Cu x O has also been estimated to support our attributions. In this case, the H 2 production rate remains approximately 184 lower (28 μmol g −1 h −1 ) than the H 2 rate obtained for 0.1% RuO 2 /1%Cu x O coupled to the 3%Ta 2 O 5 /SrZrO 3 heterostructure (5164 μmol g −1 h −1 ). The experiments indicate that charge transfer through the different heterostructure components is improved by adding 0.1%RuO 2 /1%Cu x O to 3% Ta 2 O 5 /SrZrO 3 . The photocatalytic activity for 0.1%RuO 2 /1% Cu x O is also attributed to the strong electronic coupling with Ta 2 O 5 /SrZrO 3 .
The H 2 production rate for the RuO 2 /1%Cu x O/3%Ta 2 O 5 / SrZrO 3 heterostructure of varied RuO 2 contents and other heterostructures of lower-order with Cu x O, RuO 2 , and Pt (Figure 8c,e) is contrasted with the O 2 production rate to demonstrate the overall water-splitting process (Figure 8d,f). The trends for the O 2 production rates in Figure 8d are compared to those in Figure 8c. The results show an O 2 to H 2 ratio of 1:2 for the RuO 2 /Cu x O/3%Ta 2 O 5 /SrZrO 3 heterostructure. 52 Similar ratios for lower-order heterostructures decorated with RuO 2 , Cu x O, and Pt cocatalysts can be seen in Figure 8e,f. Among the results, it should be noted that the O 2 production rate for the 0.1%RuO 2 /1%Cu x O/3%Ta 2 O 5 / SrZrO 3 prevails as the highest without evident chemical changes after reaction (Figures S5 and S6, and Table S1). Overall, the results suggest the favorable effect of the cocatalyst and bicatalyst on promoting the kinetics of O 2 evolution. 52 Although the 0.1%RuO 2 :1%Cu x O/3%Ta 2 O 5 /SrZrO 3 heterostructure prevails the highest, it is essential to reflect on the  (Figure 8e). In this case, various loadings have been assessed (i.e., 0.1, 0.3, 0.5, 1, 1.3, and 1.5 wt %) for the three RuO 2, Cu x O, and Pt cocatalysts, as shown in Figure 8e. We compare 3%Ta 2 O 5 / SrZrO 3 (1297 μmol g −1 h −1 ) with 1 wt % RuO 2 . A nearly 3fold increase (4986 μmol g −1 h −1 ) is achieved. As for the catalyst with 1 wt % Cu x O, a 2-fold increase (3282 μmol g −1 h −1 ) has been found. For Pt, a very low loading of ca. 0.1 wt % is required to obtain an activity close to 2744 μmol g −1 h −1 , which is comparable to that of either 0.5 wt % Cu x O or 0.5 wt % RuO 2 . However, the H 2 evolution activity of Pt decreases substantially comparable to that of catalysts with RuO 2 and Cu x O loadings (i.e., 0.1 wt %). In all cases, a high cocatalyst content does not necessarily improve the production of H 2 due to parasitic recombination losses as the amount of either Cu x O increases, that is, (>1 wt %), RuO 2 (>1 wt %), or Pt (>0.1 wt %). 53 Additionally, high loadings can also promote the formation of large metal (metal oxide) particles or aggregates detrimental to the overall catalytic activity during water splitting. 22 Overall, the photocatalytic activity of 0.1 wt % Cu x O and 0.1 wt % RuO 2 can be attributed to the strong electronic coupling with Ta 2 O 5 /SrZrO 3 , where hole−electron recombination might be reduced. To support our attribution, the photocatalytic activity of RuO 2 and Cu x O has been measured. The H 2 production rate for RuO 2 and Cu x O remains low, ca. 14 and 26 μmol g −1 h −1 . This indicates that Ta 2 O 5 /SrZrO 3 provides the necessary transfer of charges to RuO 2 or Cu x O, reaching the solid−liquid interface to promote H 2 water splitting. To this end, an important aspect to highlight is the reduction of the use of noble catalysts such as Ru or Pt without compromising photocatalytic activity. Even if Ru is a less costly catalyst than Pt, 28 Ru usage can be reduced when combined with other catalysts, such as Cu x O. Therefore, the photocatalytic performance of binary cocatalysts composed of RuO 2 /Cu x O has also been assessed. Various RuO 2 loadings, that is, 0.01 wt % (0.01%RuO 2 ) and 1 wt % (1%RuO 2 ), are incorporated to 1%Cu x O/Ta 2 O 5 /SrZrO 3 (Figure 8c,d).
The QE at λ = 420 nm and the photocatalysts' STH are calculated according to eqs 1 and 2 to compare our heterostructures with other systems. 54 The efficiencies obtained are summarized in Table 2 (Table S2). For example, this is the case of the SrTiO 3 -based photocatalyst with a QE of 30% at λ = 360 nm. 57 Compared to bare and decorated SrZrO 3 with Ni, Cu, Fe, and Co, our 0.1%RuO 2 /1%Cu x O/3%Ta 2 O 5 / SrZrO 3 heterostructure surpasses the known STH values by nearly 4-fold. 8 After assessing the overall water-splitting performance of the heterostructures, it is clear that the 0.1%RuO 2 /1%Cu x O/3% Ta 2 O 5 /SrZrO 3 composition has the highest H 2 or O 2 production rate and STH. Regarding QE, 0.1%RuO 2 :1% Cu x O/3%Ta 2 O 5 /SrZrO 3 has the highest among the synthesized SrZrO 3 heterostructures. 8 The QE (  Table S2. The next step is to understand the effect of the heterostructure component during charge transfer to provide a plausible picture of the water-splitting mechanism. where C is the differential capacitance, ε r is the dielectric constant of SrZrO 3 (e = 60), 62 ε 0 is the vacuum permittivity, e is the electron charge, N d is the donor density, A is the active electrode area, V is the applied potential, V FB is the flat band potential, T is the temperature (in kelvin), and k B is the Boltzmann constant. The donor density is estimated using the Mott−Schottky plot slope, and a value of 60 is estimated for the dielectric constant of SrZrO 3 . These values are summarized in Table 3.

Donor Density and Charge Transfer Resistance in the RuO 2 /Cu
In Table 3 and Figure 9a, the donor density of SrZrO 3 is affected by the incorporation of Ta Figure 8 and indicates that the photoactivity of 3%Ta 2 O 5 /SrZrO 3 can be tuned using 0.1%RuO 2 /1%Cu x O. Donor density mobility in the 0.1%RuO 2 /1%Cu x O/3%Ta 2 O 5 /SrZrO 3 heterostructure can be associated with a reduction in charge recombination ( Figure 5).
To this end, transient photocurrent measurements are evaluated under simulated solar light (100 mW cm −2 ) to understand the photocatalyst response in Figure 9b. 0.1% RuO 2 /1%Cu x O/3%Ta 2 O 5 /SrZrO 3 promoted the higher pho- toresponse associated with charge carrier separation in this heterostructure. This higher photocurrent is also attributed to the increase in light absorption. Light absorption around 250 nm or higher is improved, as shown in Figure 5c. Hence, one can assume that photogeneration of electrons and holes occurs more efficiently at the 0.1%RuO 2 :1%Cu x O/3%Ta 2 O 5 /SrZrO 3 interface than in other photocatalysts, as shown in Table 2.
For insights into the reaction kinetics, impedance analyses are carried out. The semicircle in the impedance spectra in the Nyquist plots ( Figure 10) shows the charge transfer resistance. The diameter of the semicircle describes the reaction kinetics.
A smaller diameter implies faster reaction kinetics. Figure 10 also shows the corresponding equivalent circuit, where Rs is the resistance associated with the electric connection, electrolyte, and substrate. R1 is the charge transference resistance in the electrode−electrolyte interface, and CPE is the constant phase element (  63 3.4. Charge Transfer Mechanism. Mott−Schottky plots are used to estimate the flat band potential ( Figure S7 and Table S3) by extrapolating the x-axis intercept of the linear plot (1/C 2 vs E). A positive slope is characteristic of n-type semiconductors, and a negative slope is representative of ptype semiconductors. Note that the Fermi level and the majority charge carrier band [CB (E CB ) for n-type and VB (E VB ) for p-type] can vary approximately ±0.1 V versus. NHE. 61,64,65 Therefore, it is safe to say that the band energy diagram is estimated using the Mott−Schottky and the     Figure S8. The results from Table S3 are used to understand the charge transfer mechanism (Figure 11).
The charge transfer mechanism in Figure 11 is proposed for the 0.1%Ru 2 O/1%Cu x O/3%Ta 2 O 5 /SrZrO 3 heterostructure to elucidate the possible charge pathways that led to high photocatalytic water splitting shown in Figure 8. It should be noted that other mechanisms might involve during charge transfer (e.g., Figure S9), but the mechanism in Figure 11 might be the most plausible one. The structural, morphological, chemical, optical, and electrochemical characterization results are used to derive our proposition ( Figure 11). In this heterostructure, electrons are transferred from tantalum-doped strontium zirconate to Ta 2 O 5 and Cu 2 O to overcome the evolution of H 2 . Meanwhile, the electrons in CuO move toward the RuO 2 CB. After that, these electrons recombine with Cu 2 O holes. RuO 2 holes are transferred to CuO, performing the O 2 evolution reaction. Holes in Ta 2 O 5 move to tantalum-doped strontium zirconate, where they carry out O 2 evolution reactions ( Figure 8). For other heterostructures, the possible mechanism is presented in Figure S8.

CONCLUSIONS
SrZrO 3 -based heterostructures of mixed oxides are synthesized. The highest H 2 production is ca. 5164 μmolg −1 h −1 for 0.1% RuO 2 /1%Cu x O/3%Ta 2 O 5 /SrZrO 3 , which is comparable if not even higher than that of SrZrO 3 and reported QE values for other perovskite heterostructures. In-depth structural analysis revealed the presence of Ta 2 O 5 in SrZrO 3 . Ta 2 O 5 has been found segregating at the surface and grain boundaries of SrZrO 3 , which improved the photocatalytic activity in SrZrO 3 . Yet, the photocatalytic activity of Ta 2 O 5 /SrZrO 3 is further improved with RuO 2 or Cu x O as a cocatalyst or RuO 2 /Cu x O as a binary catalyst. An optimum activity for the RuO 2 /Cu x O heterostructure components has been found, surpassing RuO 2 or Pt activity. DFT, structural, optical, and electrochemical characterization generates insights on band gap tunability for the different heterostructure components and demonstrates enhanced charge transfer for RuO 2 /Cu x O/Ta 2 O 5 /SrZrO 3 . The results are valuable in demonstrating that SrZrO 3 -based heterostructure can harvest visible light to improve the hydrogen evolution reaction. ■ ASSOCIATED CONTENT
Materials and methods for the synthesis of photocatalysts; analysis conditions employed in the characterization techniques of the physicochemical properties of the materials (XRD, SEM, TEM, XPS, UV−vis, photoluminescence, and EIS); DFT calculations; calculation details of the energy band diagram; characterization of the best photocatalyst after reaction; and supplementary microscopy images of the materials (PDF)