Unveiling Valence State-Dependent Photocatalytic Water Splitting Activity and Photocathodic Behavior in Visible Light-Active Iridium-Doped BaTiO3

Despite having favorable energetics and tunable optoelectronic properties, utilization of BaTiO3 (BTO) for photocatalytic reactions is limited by its absorption only in the ultraviolet region. To address this challenge, BTO is doped with iridium (Ir) to induce visible light absorption. The visible light-induced photocatalytic H2 generation efficiency is enhanced by 2 orders of magnitude on selective conversion of the Ir valence state from Ir4+ to Ir3+. To understand such intriguing behavior, valence state-dependent changes in the optoelectronic, structural, and surface properties and electronic band structure are comprehensively investigated. The effect of electron occupancy change between Ir4+ (t2g5 eg0) and Ir3+ (t2g6 eg0) and their energetic positions within the band gap is found to significantly influence H2 generation. Besides this, converting Ir4+ to Ir3+ enhanced the photocathodic current and lowered the onset potential. Results aid in designing photocatalysts to efficiently use low-energy photons for enhancing solar H2 production in these emerging BTO-based photocatalysts. Collectively, the observations made in this work highlight the promising application of Ir3+:BTO in z-scheme photocatalysis.


INTRODUCTION
−3 Recent demonstration of H 2 evolution via water splitting using 100 m 2 photocatalyst modules comprising aluminum-doped SrTiO 3 further demonstrated the potential of ABO 3 -type perovskites toward cost-effective and scalable solar fuel generation. 4Like SrTiO 3 , BaTiO 3 (BTO) is another promising candidate to drive photocatalytic water splitting reactions.−7 Despite these promising prospects and advantages, BTO can absorb only ultraviolet (UV) light due to its wide band gap (3.2 eV), which constitutes less than 5% of the solar spectrum. 8,9Considering the absorption cross-section and band gap of BTO, the theoretically possible maximum solar-tohydrogen (STH) energy conversion efficiency would not exceed 2%. 10 However, at least 10% STH energy conversion efficiency is required to make photocatalytic technology feasible for industrial applications.In this direction, extending the spectral response of BTO to enable maximum utilization of incoming sunlight is crucial.−16 However, the observed photocatalytic H 2 evolution rate for TM-doped perovskites could not always be correlated to the extent or magnitude of visible light absorption efficiency.For example, SrTiO 3 doped with Cr, Rh, and Sb often showed inadequate H 2 evolution compared to its undoped counterpart despite absorbing visible light. 12,14,17,18hus, merely ensuring an extended absorption via doping does not necessarily enhance the photocatalytic H 2 generation efficiency.
A careful examination of earlier reports revealed the key role of the dopant's valence state and its electronic properties in determining the H 2 generation efficiency.The valence state of Cr in Cr-doped SrTiO 3 was primarily found to be 6+. 14owever, converting Cr 6+ to Cr 3+ enhanced the photocatalytic H 2 generation activity by six times. 17−20 Converting Rh 4+ to Rh 3+ yielded factor three enhancement of photocatalytic H 2 generation activity and also exhibited photoresponse up to λ ≤ 540 nm. 21o further understand the mechanism, the photoelectrochemical performance between Rh 4+ :SrTiO 3 and Rh 3+ :SrTiO 3 is compared. 22Rh 4+ :SrTiO 3 showed a higher photocathodic current and a higher onset potential under visible light compared to Rh 3+ :SrTiO 3 .On the other hand, the valence state-dependent photoelectrochemical behavior of Ir-doped SrTiO 3 was qualitatively different compared to Rh-doped SrTiO 3 .Ir 4+ :SrTiO 3 showed a high photoanodic current under visible light, while Ir 3+ :SrTiO 3 was inert. 23The observations discussed so far highlight the complexity in understanding valence state-dependent photocatalytic activity.A comprehensive correlation among the electronic nature of the dopant, its occupancy, and its energetic position within the forbidden region of the band gap with visible light-induced H 2 generation efficiency is missing.
Among various metal dopants utilized, the choice of Ir can be understood by comparing its properties with those of wellstudied Rh-doped SrTiO 3 .With Ir being a 5d element (Rh is 4d) in the periodic table, Ir-related impurity levels are located ∼0.5 eV above that of Rh levels within the band gap of SrTiO 3 . 23Consequently, Ir-doped SrTiO 3 is expected to have optical absorption over a wider part of the solar spectrum, compared to Rh-doped SrTiO 3 .Indeed, pioneering work from the Kudo group realized extended optical absorption (λ ≤ 700 nm) in SrTiO 3 and NaTaO 3 upon Ir doping. 11,24The same group also investigated different methods of treating surfaceloaded Ir (metallic) cocatalysts and their impact on the photocatalytic H 2 evolution efficiency in Ir-doped SrTiO 3 . 11However, the effect of changing the valence state of Ir on the optoelectronic properties and/or photocatalytic activity is yet to be studied in detail.Thus, a further understanding of the intriguing effects of the Ir valence state on the photocatalytic performance would aid in the effective utilization of low-energy photons for H 2 generation, which is still a challenge.
Most of the earlier studies on dopant-induced photocatalytic activity are conducted on SrTiO 3 .Despite the advantages and unique prospects of BTO, a few reports on visible lightinduced photocatalytic activity in BTO are reported. 8,25,26Rh doping-induced p-type behavior in BTO was reported by Maeda et al. almost a decade ago. 8Recently, Shi et al. investigated the effect of the Rh doping level on the Fermi level position and n-to p-type transition was studied by isolating the cathodic contribution by the Rh dopant. 27Despite years of research related to BTO as a H 2 evolution photocatalyst, there exists no report investigating the effect of the dopant's valence state on the optoelectronic properties and subsequent photocatalytic and/or photoelectrochemical behavior of BTO.
In this direction, Ir-doped BTO as a promising visible lightabsorbing p-type material has been recently developed for the first time. 16The current work reveals valence state-dependent differences in the optoelectronic/photophysical properties and their influence on the photoelectrochemical behavior and photocatalytic H 2 evolution efficiency in Ir-doped BTO.Detailed analysis using a range of complementary tools offers insights into photocatalyst design for effectively harvesting lowenergy photons to generate H 2 .

MATERIALS AND METHODS
2.1.Photocatalyst Synthesis.Barium nitrate (99%, Merck) and titanium(IV) oxide nanopowder (NanoArc.anatase 99%, Alfa Aesar) and iridium(IV) oxide (99.9%,Aldrich) were used for synthesis without any further treatment.2 mol % Ir-doped BTO is synthesized according to the earlier report and was named as Ir 4+ :BTO. 16As obtained, Ir 4+ :BTO was subjected to heat treatment under a reducing atmosphere and is referred to as Ir 3+ :BTO (detailed methodology provided below).For comparison, undoped BTO was also synthesized according to the earlier report. 16r 4+ :BTO was reduced by a temperature-programmed reduction (TPR) technique under a H 2 flow using a Belcat II (BEL Japan) instrument.Prior to the reduction process, the sample was treated in an Ar gas flow (50 mL min −1 ) at 300 °C, followed by cooling to 50 °C.Next, the sample was subjected to a reducing environment with 5% H 2 in He at a flow rate of 30 mL min −1 , which was ramped to 400 °C at 10 °C min −1 rate and held at this temperature for 1 h.

Photodeposition of the Pt Cocatalyst and Photocatalytic H 2 Evolution
Reaction.The H 2 evolution experiments were performed in a slurry-type reactor illuminated from a side by using a 200 W Hg lamp (Superlite I 05, Lumatec), which acted as a source for both visible light (λ = 400−700 nm, total power of 500 mW) and UV light (λ = 240−400 nm, total power of 175 mW) experiments.The spectral distribution of the Hg lamp is shown in Figure S1.The experiments were carried out in a batch-type mode with the reaction solution kept at 15 °C under constant stirring at 500 rpm throughout the reaction time.In a single run, 100 mg of the photocatalyst was added to 40 mL of 10 vol % aqueous methanol solution and ultrasonicated for 5 min to obtain a homogeneous suspension.The suspension was then transferred to the reactor and purged for 10 min with argon gas at 30 mL min −1 to remove dissolved oxygen from the reaction solution.Before closing the reactor, the amount of aqueous H 2 PtCl 6 solution corresponding to the 1 wt % of Pt with respect to the photocatalyst amount was added as a precursor for in situ photodeposition of Pt, which acted as a H 2 evolution reaction cocatalyst.During light irradiation, the reactor was kept airtight.After 24 h of illumination, 200 μL of the gaseous sample was taken from the headspace of the reactor and analyzed using gas chromatography (Shimadzu GC-2030 equipped with a barrier discharge ionization detector).A 6-point calibration profile was used to accurately quantify the amount of H 2 evolved and translate it to mole values presented in the main text.

Photoelectrochemical Measurements.
Electrodes for the photoelectrochemical measurements were prepared as follows.2.5 mg of photocatalyst (Ir 4+ :BTO and Ir 3+ :BTO) was added to the solution containing 120 μL of distilled water, 100 μL of isopropanol, 5 μL of dimethylformamide, and 5 μL of Nafion.The solution was sonicated for 45 min to prepare the photocatalyst ink.20 μL of the prepared ink was drop-cast on the ITO (indium tin oxide)-coated glass substrate (resistance of <6 ohm/sq).The apparent area of the electrode containing the ink was 0.25 cm 2 area.The electrodes were dried under an infrared lamp and calcined at 300 °C for 3 h.A potentiostat (Metrohm Autolab PGSTAT 204) with a three-electrode system with platinum as the counter electrode, reversible hydrogen electrode (RHE) as the reference electrode, and ITO-coated glass substrates as the working electrode was used for photoelectrochemical measurements.0.1 M K 2 SO 4 was used as an electrolyte.The current−voltage response was recorded at the scan rate of 25 mV s −1 , and the chronoamperometric measurements were recorded at −0.2 V vs RHE.A solar simulator (Holmarc, HR-SS300WRM1-100A) equipped with an air mass 1.5 filter and a 300 W xenon short arc lamp (Ushio Inc. Japan) was the light source.The light coming from the xenon arc lamp was passed through a 420 nm long pass filter before illumination of the sample.
2.4.Structural, Surface, and Optoelectronic Characterization Tools.The powder X-ray diffraction (XRD) was performed using a Bruker D 2 Phaser using Cu Kα radiation as the radiation source of wavelength λ = 1.54 Å.The samples were mounted on the sample holder, and the XRD pattern was recorded from 20 to 90°2θ with 0.02°2θ step size and a scan speed of 0.3 s per step.
The optical absorption of the samples was measured using a UV/ vis/NIR (near-infrared) spectrometer (PerkinElmer lambda 750) in diffuse reflectance mode, equipped with an integrating sphere.
Valence states of the constituting elements of the material synthesized were investigated by X-ray photoelectron spectroscopy (XPS) (Thermo-Scientific NEXSA) using Al Kα (1486.6 eV) as the X-ray source.The operating pressure was in the ultrahigh-vacuum range from 10 −8 to 10 −10 mbar.The XPS survey spectra were recorded at a pass energy of 200 eV with a step size of 1 eV.The XPS higher energy resolution spectra were recorded with a pass energy of 50 eV with a step size of 0.1 eV.A flood gun was used to eliminate the charging effects.The spectral charge correction was performed using carbon with a C 1s peak appearing at 284.8 eV.Ultraviolet photoelectron spectroscopy (UPS) or valence band-XPS (VB-XPS) was performed using a spectrometer (Thermo-Scientific NEXSA) with a He I (21.22 eV) excitation source.The pass energy used was 2 eV with 0.050 eV step size.The operating pressure was in the ultrahigh-vacuum range from 10 −8 to 10 −10 mbar.The VB spectra of Ir 4+ :BTO and Ir 3+ :BTO were plotted by converting the counts versus kinetic energy (eV) to counts versus binding energy (eV).Kinetic energy to binding energy conversion was performed using the equation, photon energy = binding energy + kinetic energy, considering that the spectrometer is calibrated such that the Fermi energy will appear at binding energy zero.Due to the sensitivity of the XPS instrument, BTO doped with 3 mol % of Ir was essential to better visualize the differences in the changes in the valence state of Ir (after H 2 -TPR experiments) in the Ir 4f peak binding energy.
The surface area of the photocatalysts was determined using a Belsorb (II) mini-instrument (BEL, Japan).The N 2 sorption analysis was performed at 77 K to measure the Brunauer−Emmet−Teller (BET) specific surface area.

Computational Method to
Analyze the Electronic Structure.Computational investigations were undertaken using density functional theory (DFT).−32 The electron wave function is expanded in plane waves with a cutoff energy of 500 eV.The cell and atomic relaxations are carried out with the energy and force convergence criteria set to 10 −6 eV and −0.01 eV/Å, respectively.For the simulation of Ir doping in BTO, a 135-atom supercell with a Monkhorst−Pack k-mesh of 3 × 3 × 3 is used to sample the Brillouin zone for geometry optimization and appropriate denser mesh is used to calculate the density of states.Note that Ir is doped at the Ti site in the BTO cell.Since band gaps are not quantified correctly by GGA, we employ the rotationally invariant approach of GGA + U, where U values of 4 and 8 eV are used for Ti and O, respectively.For Ir, a U value of 2 eV is used in the present work.The Ir valence state-dependent optical absorption spectra are calculated by converting the complex dielectric function obtained from the VASP to the absorption coefficient.To determine the optical response of Ir-doped BTO, the frequency-dependent complex dielectric function ε(ω) = ε 1 (ω) + iε 2 (ω) has been calculated, which indicates the linear response of material to the electromagnetic field.The electronic structure and dielectric function are related to each other.Here, the imaginary part of dielectric function ε 2 (ω) is given by the summation over the large number of empty states and the real part ε 1 (ω) is obtained from ε 2 (ω) using Kramers−Kronig relation.Since the absorption range of the solar spectrum is crucial in optoelectronic materials, the optical absorption coefficient α(ω) is calculated using the following relation The spectra of absorption coefficient vs wavelength are obtained by taking the average of three principal non-zero diagonal constituents (α xx (ω), α yy (ω), and α zz (ω) corresponding to the a, b, and c axes, respectively).

RESULTS AND DISCUSSION
The first part will discuss the controlled transformation of the Ir valence state from Ir 4+ to Ir 3+ and the resulting optoelectronic/structural/surface properties.The latter part will focus on elucidating valence state-dependent photoelectrochemical behavior and photocatalytic H 2 evolution activity.
Typically, the valence state of the dopant in ABO 3 -type titanate-based perovskites is manipulated by heating the sample under reducing conditions, for example, in the presence of H 2 gas. 11,17,33In this approach, obtaining precise information on the selectivity toward a given reducible species is rather complex, i.e., it is difficult to find out whether the constituents of the perovskite host or the dopant is reduced.Hence, in the current work, the valence state of the Ir dopant from Ir 4+ to Ir 3+ is selectively converted by performing the H 2 -TPR experiment.In this method, the as-prepared Ir 4+ :BTO is subjected to a linear heating process in an enclosed furnace with a H 2 flow.The difference in the reducing gas (H 2 ) concentration before and after the reduction reaction is monitored by thermal conductivity detector (TCD) signal vs temperature.The H 2 -TPR experiment offers control/selectivity toward the reduction reaction and provides unique information to identify the chemical/electronic nature of the species reduced, unlike typically employed uncontrolled heating of the sample. 34igure 1 compares the H 2 -TPR profiles of Ir 4+ :BTO and undoped BTO.The presence of a distinct TCD signal response for Ir 4+ :BTO, which is virtually absent for undoped BTO, indicates that Ir-related species are being reduced during the H 2 -TPR experiment.Such comparison allowed us to rule out the possibility of the H 2 -TPR experiment reducing the constituents of host BTO lattice.Furthermore, the magnitude of the TCD signal (Figure S2) concomitantly decreased for samples with a lower Ir doping level.To further elucidate the electronic nature of the reduced Ir-related species, the H 2 -TPR experiment with IrO 2 was conducted (Figure S3).Note that IrO 2 was employed as the precursor to synthesize Ir-doped BTO that yielded a high concentration of Ir 4+ species. 16On comparing H 2 -TPR profiles of Ir 4+ :BTO and IrO 2 , a similarity in the temperature range is noticed.Thus, the appearance of the peak in the H 2 -TPR profile of Ir 4+ :BTO is due to the reduction of Ir 4+ to its lower valence state.
To confirm the valence state of Ir 4+ after the H 2 -TPR experiment, an XPS analysis is conducted.Figure 2 collectively presents the Ir 4f core level spectra of as-prepared Ir 4+ :BTO (before the H 2 -TPR experiment) and Ir 3+ :BTO obtained after the H 2 -TPR measurement.Figure 2a shows ≈1.3 eV shift toward lower binding energy for Ir 3+ :BTO compared to Ir 4+ :BTO.However, the binding energy shift for Ti 2p and Ba 3d is <0.1 eV (Figure S4), suggesting a high selectivity toward Ir-related reducible species during the H 2 -TPR experiment.This observation also agrees with the observation of a virtually absent TCD signal for undoped BTO in Figure 1.
Figures 2b and 2c depict the deconvoluted Ir 4f core level spectra of Ir 4+ :BTO before and after the H 2 -TPR experiment, respectively.The Ir 4f 7/2 peak at ≈61.5 eV indicated in purple is characteristic of Ir in the 3+ valence state, while the peak at ≈62.5 eV indicated in orange corresponds to Ir in the 4+ valence state. 23Ir 4+ :BTO showed a high percentage (66%) of Ir 4+ .The concentration of Ir 3+ increased from 33 to 91% after the H 2 TPR experiment, whereas the Ir 4+ content concomitantly decreased from 66 to 9%, and this sample is referred to as Ir 3+ :BTO.Such a significant increase in Ir 3+ concentration is attributed to the selective conversion of Ir 4+ to Ir 3+ during the H 2 -TPR experiment.It is essential to note the absence of an XPS peak around 60.8 eV corresponding to Ir 0 (metallic Ir) 35 after H 2 -TPR experiments (Figure 2c).
XPS can primarily probe the changes in the valence state of the elements and dopants located on the surface.Note that the reduction method using the H 2 -TPR experiment is expected to induce changes in the bulk of the photocatalyst as well.To understand such a behavior, XRD and optical properties are investigated.Figure 3 compares the XRD patterns of Ir 4+ :BTO and Ir 3+ :BTO.By comparing with the database corresponding to undoped BTO (ICSD 27970), the cubic phase is retained for Ir 3+ :BTO, and the overall structural integrity is maintained.As discussed earlier, the H 2 -TPR experiment successfully converted a majority of the Ir 4+ in the as-synthesized Ir-doped BTO sample to Ir 3+ .Note that Ir 3+ (82 pm) has a greater ionic radius than does Ir 4+ (76.5 pm).Considering the Ir dopant occupying Ti sites of the BTO host lattice, 16 the conversion of Ir 4+ to Ir 3+ is expected to result in a lattice expansion.As expected, a shift toward a lower diffraction angle by ≈0.27°2θ for Ir 3+ :BTO compared to Ir 4+ :BTO is observed.This observation indicates that the Ir valence state change after H 2 -TPR is not just on the surface but rather a bulk phenomenon.
Understanding the valence state-dependent optical properties offers insights into the electron occupancy and energetic position of the dopants in the forbidden region of the band gap. Figure 4a presents the optical absorption spectra of undoped BTO, Ir 4+ :BTO, and Ir 3+ :BTO.Besides the fundamental transition (T 1 in Figure 4d) from the O 2p orbitals of VB to Ti 3d of CB, a broad absorption from 390 to 600 nm noticed for Ir 4+ :BTO (T 2 in Figure 4d) is attributed to the electron transitions from VB to Ir 4+ (Ir 5d t 2g 5 e g 0 configuration) in-gap energy levels.Due to the partially occupied d 5 electronic configuration of Ir 4+ , optical transitions from Ir 4+ to the CB are unlikely.Let us now focus on the optical absorption of Ir 3+ :BTO, which is clearly found to be further red-shifted to 720 nm compared to Ir 4+ :BTO.The extended tail absorption from around 600 to 720 nm can be attributed to the electron transitions from Ir 3+ (Ir 5d t 2g 6 e g 0 configuration) to the CB (T 3 in Figure 4d).Thus, for Ir 3+ : BTO, visible light photons can populate electrons in the CB, which is expected to improve the photocatalytic H 2 evolution.The effect of the Ir valence state on the optical absorption property is further investigated via the computational approach.Figure 4b depicts the absorption coefficient vs wavelength plot of undoped BTO, Ir 4+ :BTO, and Ir 3+ :BTO.Ir 4+ :BTO is obtained by substituting Ir for Ti sites of the BTO host. 16By adding one more electron to Ir 4+ , Ir 3+ in BTO is modelled for the analysis.Ir-doped BTO photocatalysts exhibit  extended optical absorption in the visible/NIR region compared to undoped BTO absorbing only in the UV region.The absorption of Ir 3+ :BTO extends beyond that of Ir 4+ :BTO toward the NIR region, owing to the pronounced T 3 transition indicated in Figure 4d.Thus, the computational analysis of the electronic structure also shows a similar trend in valence statedependent optical absorption properties and is in qualitative agreement with Figure 4a.
Figure 4c compares the VB spectra of Ir 4+ :BTO and Ir 3+ :BTO obtained from UPS.This experiment aims to understand how a change in the Ir valence state impacts the electronic band structure.A change in the electron occupancy upon conversion of Ir 4+ to Ir 3+ is expected to influence the density of states (DOS) and the Fermi level (E F ) position.Indeed, the observed ≈0.9 eV shift toward higher binding energy corresponding to the spectral onset for Ir 3+ :BTO compared to Ir 4+ :BTO indicates an upward shift in the E F position for Ir 3+ :BTO.The overall shape of the VB spectra changes, suggesting an increase in the DOS of Ir 3+ states for Ir 3+ :BTO compared to Ir 4+ :BTO.Conclusions from Figure 4c agree well with the valence state-dependent electronic structure analysis (Figure S5) elucidated using the computational approach.
Figure 4d presents a band energy diagram based on the data discussed earlier.The VB maximum for most of the metal oxides is considered to be situated approximately at +3 eV vs the normal hydrogen electrode (NHE). 36Considering the band gap of undoped BTO as 3.2 eV, the absorption onsets of Ir 4+ :BTO and Ir 3+ :BTO, and the E F position derived from the VB spectra, the band energy diagram is deduced.By virtue of its occupancy, Ir 4+ levels located in the in-gap region can potentially act as electron traps.Considering the energetic position of Ir 4+ levels with the potential required to reduce the H + of water, an inadequate thermodynamic driving energy to generate H 2 can be anticipated.However, after the H 2 -TPR experiment, the density of Ir 3+ levels is increased, thus promoting the T 3 transition.As a result, electrons can populate the CB, and they have sufficient energy to drive the H 2 generation.Therefore, the valence state of the Ir dopant is found to significantly impact optical, electronic, structural, and surface properties.The following part of the Results and Discussion will reveal valence state-dependent photoelectrochemical behavior and photocatalytic H 2 evolution activity.

Effect of the Ir Valence State on the Photoelectrochemical Behavior. As discussed earlier, converting
Ir 4+ to Ir 3+ in Ir-doped BTO has a significant influence on the optical absorption and the electronic band structure.To further probe these phenomena, the effect of the Ir valence state on the photoelectrochemical behavior is investigated by conducting linear sweep voltammetry, and the corresponding voltammograms (LSV) are presented in Figure 5. Figures 5a  and 5b depict LSV of Ir 4+ :BTO and Ir 3+ :BTO, respectively, recorded in the dark and under visible light (λ > 420 nm) irradiation.For comparison, the LSV of bare ITO performed in the dark is provided in Figure 5a and shows a negligible response.Comparing the respective LSVs, photoinduced enhancement in the current is higher (light vs dark) for Ir 3+: BTO compared to Ir 4+ :BTO.At −0.2 V vs RHE; Ir 3+ :BTO shows a factor of ≈1.7 increment in the photocurrent than Ir 4+ :BTO.Similar differences in valence state-dependent photoelectrochemical behavior are observed for Ir-doped BTO with a higher doping level as well (Figure S6).Furthermore, visible light-induced enhancement in the cathodic current is supported by chronoamperometric studies performed at −0.2 V vs RHE, depicted in Figure 5c.Observing a negative photocurrent for both Ir 3+ :BTO and Ir 4+ :BTO is in accordance with the LSVs discussed earlier.Moreover, both the photocatalysts exhibit a nondecaying photocurrent response in the time window measured, highlighting their stability during the visible light irradiation.A factor of ≈1.6 increment in the photocurrent observed for Ir 3+ :BTO compared to Ir 4+ :BTO indicates its utility in photocatalytic reactions.The observed increment in photocurrent during the chronoamperometry is further validated by monitoring the H 2 evolved as illustrated in Figure S7.An increment of ≈2.5 times in H 2 evolution recorded for Ir 3+ :BTO with concurrent enhancement in Faradaic efficiency (calculated at the end of 4 h, refer the Supporting Information) by a factor of ≈2.3 times relative to Ir 4+ :BTO clearly demonstrates the Ir valence state effect on the photoelectrochemical activity.Figure 5c compares the photocurrent onset between Ir 3+ :BTO and Ir 4+ :BTO recorded under intermittent visible light irradiation.A minimum of ≈0.2 V shift in the onset potential toward a lower cathodic side is noticed for Ir 3+ :BTO compared to Ir 4+ :BTO.A similar valence state-dependent shift in the onset potential was observed in Rh-doped SrTiO 3 . 22n short, the following information can be deduced through a collective analysis of Figure 5.A higher photocurrent magnitude with enhanced Faradaic efficiency accompanied by a lower onset potential corroborates the positive effect of extended light absorption and electron occupancy of the Ir dopant within the forbidden band gap region of Ir 3+ :BTO.These observations further aided in establishing the role of the Ir valence state in the optoelectronic properties and electronic band structure; both favor the applicability of Ir 3+ :BTO for the H 2 evolution reaction.Hence, the visible light-induced enhancement in cathodic response paired with lower onset potential highlights the promising application of the Ir 3+ :BTO photocathodic material toward z-scheme photocatalysis.

Effect of the Ir Valence
State on the Photocatalytic H 2 Evolution Activity.Figure 6 depicts the effect of the Ir valence state on the photocatalytic H 2 evolution activity under visible light (λ > 400 nm).Ir 3+ :BTO shows 2 orders of magnitude enhancement in H 2 evolution compared to Ir 4+ :BTO.Note that it is to rule out that the H 2 -TPR measurement that converted Ir 4+ to Ir 3+ did not drastically alter the surface area.The BET surface area measurements (Table S1) after H 2 -TPR showed an increment from 3 to 3.5 m 2 g −1 for Ir 3+ :BTO compared to Ir 4+ :BTO.This minor increment (factor 1.2) in the surface area cannot explain 2 orders of magnitude increment in the H 2 evolution activity observed for Ir 3+ :BTO compared to Ir 4+ :BTO.Thus, the data presented in Figure 6 can indeed be attributed to the changes in the valence state of Ir, which is explained as follows.The partially occupied Ir 4+ in Ir 4+ :BTO acts as deep traps (T 2 transition in Figure 4d) for electrons and thus is expected to reduce the electron lifetime.Consequently, electron transport toward the water interface for H 2 generation will be inefficient.A similar observation was noticed in the Rh 4+ -doped SrTiO 3 . 19o further study this behavior in Ir-doped BTO, an electrochemical impedance spectroscopy analysis was conducted.Electrochemical impedance spectra in Figure S8 illustrate a larger semicircle for Ir 4+ :BTO, which is indicative of higher charge transfer resistance compared to Ir 3+ :BTO.This would further impede electron transport towards the surface of the photocatalyst to realize H 2 evolution in Ir 4+ :BTO.This notion is supported by observing lower photocurrent density for Ir 4+ :BTO discussed earlier in Figure 5.Moreover, the energetic position of Ir 4+ is thermodynamically not favorable to realize H 2 generation.All of these reasons collectively lead to H 2 evolution of 0.008 μmol for Ir 4+ :BTO under visible light (λ > 400 nm).
Ir 3+ :BTO showed 2 orders (0.81 μmol) of magnitude enhancement in H 2 generation than Ir 4+ :BTO, which is rationalized as follows.For Ir 3+ :BTO, the concentration of electron-trapping Ir 4+ levels is decreased, and concurrently, the T 3 transition (Figure 4d) is pronounced that promotes electrons to the CB instead of trapping to Ir 4+ levels.Both of these effects are expected to collectively contribute in extending the electron lifetime for Ir 3+ :BTO compared to Ir 4+ :BTO, which consequently favors electron transfer to enhance the H 2 generation yield.A similar trend in valence state-dependent H 2 evolution activity was observed (Tables S2  and S3) when measured under UV light and for Ir-doped BTO with a higher doping level.These results unambiguously demonstrate the crucial role of the Ir valence state in realizing pronounced visible light-induced H 2 generation.
In the work presented here, the emphasis is to reveal the origin of valence state-dependent optoelectronic properties, photoelectrochemical response, and photocatalytic H 2 evolution generation efficiency.We are currently in the process of identifying parameters to enhance the H 2 evolution activity further, particularly in λ > 500 nm.Furthermore, we are planning to construct a z-scheme photocatalyst with this photocathode material.Besides photocatalytic applications, realizing optical control over the Ir valence state may also find application in other fields such as spintronics and optoelectronics.

CONCLUSIONS
The H 2 -TPR experiment selectively and efficiently converted Ir 4+ to Ir 3+ in Ir-doped BTO photocatalysts.The origin of Ir valence state-dependent photocatalytic H 2 evolution efficiency was revealed using a range of complementary spectroscopy tools and computational analysis of the electronic structure.The Ir 3+ :BTO, which showed an increment of 2 orders in photocatalytic H 2 generation and a factor of ≈2.3 enhancement in Faradaic efficiency recorded during chronoamperometry relative to Ir 4+ :BTO, had filled in-gap Ir 3+ donor levels, allowing optical transition to the CB, populating the CB with free electrons.In contrast, partially occupied Ir 4+ levels in Ir 4+ :BTO acted as a sink for photogenerated electrons, thus reducing the number of free electrons in the CB available for H 2 evolution.This work established a comprehensive correlation among the energetic position of the dopant, its electron occupancy, the resulting optoelectronic properties, and photocatalytic H 2 evolution activity.These results help in designing photocatalysts to efficiently harness low-energy visible/NIR photons to realize enhanced photocatalytic activity.Collectively, the promising features of Ir 3+ :BTO revealed in this work demonstrate its potential as a photocathode for visible light-induced z-scheme reactions.