A Z-scheme Ba₂AgIO₆/C₃N₄ heterojunction enabling efficient visible-light photocatalytic H₂O₂ production via the direct one-step two-electron O₂ reduction reaction†

Kexin Yang‡ ^a, Zhijia Song‡^a, Xiaoxiao Fu^a, Wei Yan^a, Tao Zhang^a, Yiwei Hua^a, Haixin Lin*^ab, Qin Kuang*^ab and Zhaoxiong Xie^ab
aState Key Laboratory of Physical Chemistry of Solid Surfaces, Collaborative Innovation Center of Chemistry for Energy Materials, Department of Chemistry, College of Chemistry and Chemical Engineering, Xiamen University, Xiamen 361005, China. E-mail: qkuang@xmu.edu.cn; haixinlin@xmu.edu.cn
^bInnovation Laboratory for Sciences and Technologies of Energy Materials of Fujian Province (IKKEM), Xiamen 361005, China

First published on 12th July 2025

---

---

1. Introduction

Hydrogen peroxide (H₂O₂) is a critical chemical compound with widespread applications in various industries, including the chemical, healthcare, energy, and environmental sectors.^1–3 Conventional industrial production of H₂O₂ primarily relies on the anthraquinone process,² which is energy-intensive and generates significant wastewater, leading to considerable environmental concerns. In contrast, photocatalytic H₂O₂ production requires only H₂O and O₂, which makes it an attractive alternative to traditional methods.^4–8

The pathways for photocatalytic H₂O₂ production consist of the oxygen reduction reaction (ORR) and water oxidation reaction (WOR). Within the ORR, there are two mechanisms: an indirect two-step 2-electron (e⁻) process (O₂ + e⁻ → ˙O₂⁻, ˙O₂⁻ + e⁻ + 2H⁺ → H₂O₂) and a direct one-step 2e⁻ process (O₂ + 2e⁻ + 2H⁺ → H₂O₂).⁹ However, ˙O₂⁻ intermediates are highly oxidizing and may lead to the degradation of certain metal-free catalysts in photosynthetic processes, thereby limiting the practical applicability of photocatalysts.¹⁰ Meanwhile, the direct one-step 2e⁻ ORR is more favorable in terms of thermodynamics, product purity and reaction rate, in comparison with the indirect two-step 2e⁻ ORR.^11–14 Therefore, it is essential to design photocatalytic H₂O₂ production by the direct one-step 2e⁻ ORR. For example, Jiao et al. demonstrated that doping Cd into S-vacancy-rich Zn₃In₂S₆ can promote the Yeager-type adsorption configuration of O₂ and optimize the direct one-step 2e⁻ ORR pathway, resulting in highly active and selective H₂O₂ generation.¹¹ Furthermore, Luo et al. achieved efficient H₂O₂ generation via the direct one-step 2e⁻ ORR pathway by introducing a sulfone unit into COFs to change the O₂ adsorption configuration from the Pauling-type to the Yeager-type.¹³

In recent years, graphitic carbon nitride (g-C₃N₄), as a metal-free semiconductor photocatalyst, has attracted significant attention due to its visible light absorption, suitable energy band structure, and high chemical stability, making it a promising candidate for photocatalytic H₂O₂ production.^15–19 However, in pure C₃N₄, the photogenerated charge carriers easily undergo rapid recombination, and consequently, most C₃N₄-based photocatalysts typically produce H₂O₂ through the indirect two-step 2e⁻ ORR.^14,20–24 Constructing heterojunctions not only promotes the separation and migration of charge carriers,^25,26 but also modulates the adsorption configuration of oxygen on the photocatalyst surface, making efficient H₂O₂ generation via the direct one-step 2e⁻ ORR pathway possible. For example, Liu et al. transformed the O₂ adsorption configuration from the Pauling-type to the Yeager-type by depositing FeOOH onto C₃N₄.¹² This transformation suppressed the formation of superoxide radicals (˙O₂⁻) and promoted H₂O₂ generation via the direct one-step 2e⁻ ORR pathway.

Compared with the traditional Type II heterojunction, the Z-scheme heterojunction shows significant advantages in photocatalytic H₂O₂ production. Firstly, due to the spatial separation of the reduction site and the oxidation site, the recombination of photogenerated carriers is effectively inhibited. Secondly, its unique Z-scheme charge transfer mechanism enables the oxidation half-reaction and the reduction half-reaction to occur respectively on semiconductors with stronger oxidation and reduction capabilities, thereby expanding the overall redox capacity and enhancing the efficiency of the reaction.^26–29 Benefitting from the above advantages, the photocatalytic H₂O₂ yield of Z-scheme Cd_0.6Zn_0.4S/ultrathin g-C₃N₄ was as high as 1098.5 μmol g⁻¹ h⁻¹, exceeding those of most CN-type and sulfide-type photocatalysts.²⁶ As a narrow-bandgap semiconductor, Ba₂AgIO₆ (BAIO) exhibits significant potential for application in visible light photocatalysis.^30,31 In this study, in situ synthesis was conducted to construct BAIO/C₃N₄ Z-scheme heterojunction photocatalysts, with BAIO and C₃N₄ serving as the ORR and WOR centers, respectively. The establishment of the Z-scheme heterojunction greatly enhanced the separation and migration of charge carriers. The incorporation of BAIO effectively induced a transition in the adsorption conformation of O₂ to the Yeager-type, leading to a more efficient and selective direct 2e⁻ ORR for H₂O₂ generation. Therefore, the optimal sample (with a BAIO:C₃N₄ ratio of 1:1.5) achieved a H₂O₂ yield of 535.9 μmol g⁻¹ h⁻¹ under visible light irradiation, demonstrating an 8.8-fold enhancement in performance compared to pure C₃N₄. This study emphasizes the crucial role of modulating the O₂ adsorption configuration and reaction pathway in optimizing photocatalytic efficiency, providing a promising route for enhancing H₂O₂ production in artificial photosynthetic systems.

2. Experimental

2.1. Chemicals

Urea (AR), acetonitrile (HPLC, AR), nitric acid (HNO₃, AR), ethyl alcohol (C₂H₅OH, AR) and anhydrous sodium sulfate (Na₂SO₄, AR) were purchased from Sinopharm Chemical Reagent Co., Ltd. Silver oxide (Ag₂O) and periodic acid (H₅IO₆) were purchased from Acmec. Barium hydroxide octahydrate (Ba(OH)₂·8H₂O) was obtained from Aladdin Reagent, Co., Ltd (Shanghai, China). All chemicals were used without further purification. Deionized water was used in all experiments.

2.2. Synthesis of photocatalysts

2.3. Characterization of materials

The crystal structures of the as-prepared photocatalysts were observed by powder X-ray diffraction (XRD, Rigaku Ultima IV). The morphological characteristics of the photocatalysts were determined by scanning electron microscopy (SEM, Hitachi S-4800) and transmission electron microscopy (TEM, Talos F200s) at an acceleration voltage of 200 kV. The elemental chemical state and valence band of the photocatalysts were analyzed by X-ray photoelectron spectroscopy (XPS, AXIS Supra+, Shimadzu). UV-vis diffuse reflectance spectroscopy (UV-vis DRS) was performed on an UV-vis photometer (Cary 5000, Agilent) with BaSO₄ as the background. Steady-state photoluminescence (PL) and time-resolved photoluminescence (TR-PL) spectra were recorded using F-7000 (Hitachi) and FLS1000 (Edinburgh) fluorescence spectrophotometers, respectively. Electron paramagnetic resonance (EPR) spectra were obtained on an EPR spectrometer (Bruker, EMX-10/12). N₂ adsorption–desorption analysis was performed by the Brunauer–Emmett–Teller (BET) method on a Micromeritics (ASAP 2460) surface area analyzer to determine the specific surface area of the photocatalysts. The zeta potential was measured using a Malvern Zetasizer (Nano ZS90) in pure water.

2.4. Photocatalytic synthesis of H₂O₂

25 mg of photocatalyst was placed in 50 mL of water and evenly dispersed by ultrasonic treatment. Subsequently, O₂ was introduced into the system for 30 minutes to saturate it. After that, the device was irradiated with a 300 W Xe lamp (150 mW cm⁻²) equipped with a >420 nm cutoff film (λ ≥ 420 nm) for photocatalytic testing. The whole reaction process was kept agitated, and the temperature was controlled at 20 °C using circulating water. Every 15 minutes, 2 mL of the solution was extracted from the device, and the photocatalyst was separated through filtration. The experiment on photocatalytic H₂O₂ generation under natural sunlight was conducted from 13:00 to 14:00 on 4 December 2024 at Xiamen University. The average light intensity was found to be 35 mW cm⁻².

In our study, the H₂O₂ content was determined by iodometry.³² To prepare the reaction mixture, 0.25 mL of the reaction solution was combined with 1.75 mL of water, 1 mL of 0.1 mol L⁻¹ potassium hydrogen phthalate (C₈H₅KO₄) solution and 1 mL of 0.4 mol L⁻¹ potassium iodide (KI) solution. The mixture was allowed to stand for 30 minutes to complete the reaction. During this time, H₂O₂ reacted with I⁻ according to the following equation: H₂O₂ + 3I⁻ + 2H⁺ → I₃⁻ + 2H₂O. The strong absorption peak of I₃⁻ can be detected at around 350 nm by UV-vis spectroscopy, and the amount of H₂O₂ produced can be calculated according to the standard working curve.

2.5. H₂O₂ decomposition study

The decomposition of H₂O₂ was conducted by mixing 20 mg of the BAIO/C₃N₄ heterojunction with 50 mL of a 2 mM H₂O₂ aqueous solution. The photocatalyst suspension was irradiated using a 300 W xenon lamp. The reaction was maintained at 20 °C using cooling water. A 1 mL dispersion liquid was taken from the suspension every 15 minutes, and filtered through a 0.22 μm syringe filter to remove the photocatalyst. The amount of H₂O₂ was determined using the iodometric method.

2.6. Photoelectrochemical measurements

The Mott–Schottky analysis, electrochemical impedance spectroscopy (EIS), and transient photocurrent response and open-circuit potential (OCP) measurements of the photocatalysts were performed on an electrochemical workstation (CHI 650E) equipped with a three-electrode system. To fabricate the working electrode, 5 mg of photocatalyst was dispersed in 1 mL of isopropyl alcohol and 20 μL of Nafion solution by ultrasonic treatment for 1 h to form an ink. A 0.1 mL photocatalyst dispersion was drop-coated onto an indium tin oxide (ITO) conductive glass and allowed to dry naturally under ambient conditions. A platinum mesh and an Ag/AgCl electrode were used as the counter electrode and the reference electrode, respectively. The electrolyte used was a 0.1 M Na₂SO₄ aqueous solution. Transient photocurrent responses were recorded under illumination from a 300 W Xe lamp.

A rotating ring disk electrode (RRDE) was used to measure the number of transferred electrons (n) of the sample in the ORR. 4 mg of photocatalyst was dispersed in 700 μL of isopropyl alcohol, 300 μL of H₂O and 10 μL of Nafion solution to form an ink, which was uniformly dispersed using ultrasound for 1 h. Next, 10 μL of ink was dropped onto the RRDE and dried in air. A graphite rod electrode served as the counter electrode, while a Hg/HgO electrode was used as the reference electrode. An O₂-sufficient 0.1 M KOH solution was used as the electrolyte. The speed of the RRDE was set to 1600 rpm and the potential range was set to 0–1.0 V vs. RHE. The number of transferred electrons (n) was calculated according to the following formula:

where I_r is the ring current, I_d is the disk current, and N₀ is the collection efficiency (N₀ = 0.37).

3. Results and discussion

3.1. Synthesis and characterization

As illustrated in Fig. 1a, the Ba₂AgIO₆/C₃N₄ heterojunction was constructed by growing Ba₂AgIO₆ (BAIO) in situ on the surface of C₃N₄. The scanning electron microscopy (SEM, Fig. S1†) and transmission electron microscopy (TEM, Fig. 1b–d) observation revealed that the BAIO sample consisted of nanoparticles with sizes ranging from 10 to 20 nm, while C₃N₄ exhibited a typical two-dimensional nanosheet morphology. In the heterojunction with a BAIO/C₃N₄ mass ratio of 1:1.5, the BAIO nanoparticles were randomly dispersed on the C₃N₄ nanosheets. Based on the Brunauer–Emmett–Teller (BET) method, the specific surface areas of BAIO, C₃N₄, and BAIO/C₃N₄ were 60.7, 212.8, and 104.3 m² g⁻¹, respectively (Fig. S2†). The specific surface area of the heterojunction sample was between those of BAIO and C₃N₄, which was consistent with expectations.

The structure of the BAIO/C₃N₄ heterojunction was studied using powder X-ray diffraction (XRD), Fourier transform infrared spectroscopy (FTIR) and Raman spectroscopy. Fig. 1e shows the XRD patterns of BAIO, C₃N₄, and Ba₂AgIO₆/C₃N₄ heterojunctions with varying mass ratios. The diffraction peaks of BAIO and C₃N₄ aligned closely with those reported in the literature.^30,33 The XRD patterns of the BAIO/C₃N₄ heterojunction samples exhibited a combination of the peaks from both components, with the intensity of the BAIO diffraction peaks gradually increasing as its content increased. The FTIR spectra of BAIO, C₃N₄, and BAIO/C₃N₄ are compared in Fig. 1f. The FTIR spectrum of BAIO/C₃N₄ was similar to that of C₃N₄. Raman spectroscopy was a more suitable technique for investigating the Ba–O–Ba vibrational modes, and the results are shown in Fig. 1g. The Raman spectra of BAIO/C₃N₄ displayed small peaks corresponding to Ba–O–Ba (674 cm⁻¹)³⁴ and characteristic peaks of C₃N₄ (710 cm⁻¹ and 1236 cm⁻¹).³⁵ These observations collectively confirmed the successful integration of BAIO and C₃N₄ in the BAIO/C₃N₄ heterojunctions.

To further investigate the interaction at the BAIO and C₃N₄ interfaces, we performed X-ray photoelectron spectroscopy (XPS) of the samples (Fig. 2). All binding energies were calibrated using the C 1s peak of contaminant carbon (284.8 eV) as the reference. In the C 1s spectrum of C₃N₄, the peak at 288.3 eV corresponded to the NC–N bond (Fig. 2a). The N 1s spectrum revealed four peaks centered at 398.8 eV, 399.7 eV, 401.1 eV, and 404.5 eV, which are attributed to CN–C, N– (C)₃, amino functionalities (N–H), and the π-excitation of C–N heterocycles, respectively (Fig. 2b).^36–38 For BAIO, the peaks at 780.2 eV and 795.7 eV were assigned to Ba 3d_5/2 and Ba 3d_3/2, while the peaks at 624.3 eV and 635.9 eV were ascribed to I 3d_5/2 and I 3d_3/2 in BAIO (Fig. 2c and d). Based on the binding energy values, it can be concluded that I had a relatively high oxidation state.³⁹ The peaks located at 374.3 eV and 368.3 eV fitted with Ag 3d_5/2 and Ag 3d_3/2, respectively (Fig. S4a†). Since both components involved oxygen, the shift in the O 1s spectrum for the heterojunction was not analyzed (Fig. S4b†).^37,40 Compared to the individual components, the C and N peaks of BAIO/C₃N₄ shifted towards lower binding energies, while the Ba, I, and Ag peaks shifted towards higher binding energies. The XPS results indicated a charge transfer from BAIO to C₃N₄ at the interfaces.

3.2. Photocatalytic H₂O₂ production performance

To evaluate the effectiveness of the heterojunction construction strategy, the photocatalytic H₂O₂ production performance of BAIO/C₃N₄ heterojunctions with different ratios under visible light was compared with BAIO and C₃N₄ (Fig. 3a). The yields of H₂O₂ were measured using iodometry, and the linear fitting formula for the standard H₂O₂ concentration is shown in Fig. S5.† The H₂O₂ yields of BAIO/C₃N₄ heterojunctions were all higher than those of BAIO and C₃N₄, exhibiting a volcano-shaped relationship with the content of BAIO. The sample with a BAIO:C₃N₄ ratio of 1:1.5, i.e., BAIO/C₃N₄, exhibited the best performance, achieving a H₂O₂ yield of up to 535.9 μmol g⁻¹ h⁻¹, which was 4.8 times and 8.8 times higher than those of BAIO and C₃N₄, respectively. Notably, the H₂O₂ yield of BAIO/C₃N₄ is at the current state of the art for C₃N₄-based photocatalysts (Table S1†), suggesting its promising potential. Fig. 3b shows that the yield of H₂O₂ generally exhibits an upward trend; it does not strictly adhere to a linear relationship over time. This was primarily due to the simultaneous occurrence of a decomposition reaction during H₂O₂ generation (Fig. S6†). In addition, the stability of the BAIO/C₃N₄ heterojunction in photocatalytic H₂O₂ production was further investigated (Fig. 3c). Clearly, the H₂O₂ yield of BAIO/C₃N₄ showed little change after five cycles, indicating its decent stability. The post-reaction BAIO/C₃N₄ was characterized using TEM, XRD (Fig. S7†) and XPS (Fig. S8†), and no obvious changes in morphology and structure were observed. The photocatalytic H₂O₂ production was further tested under outdoor conditions using natural sunlight, maintaining a yield of 276.3 μmol g⁻¹ h⁻¹. This highlights its potential for large-scale H₂O₂ production (Fig. 3d).

3.3. Band structure and photogenerated charge transfer

To investigate the reasons for the performance discrepancy, the light absorption ranges of the photocatalysts were assessed using UV-vis diffuse reflectance spectroscopy (UV-vis DRS) (Fig. 4a). The results showed that the introduction of BAIO significantly enhanced the heterojunction's light absorption capacity compared to the pure C₃N₄. The corresponding Tauc plots (Fig. 4b) confirmed the band gap (E_g) values of 2.07 eV for BAIO and 3.01 eV for C₃N₄, respectively.^41–43 According to Mott–Schottky measurements (Fig. 4c and d), the flat-band potentials (E_fb) of BAIO and C₃N₄ were found to be −0.80 V and −0.54 V (vs. NHE, pH = 0). In addition, the positive slopes of the Mott–Schottky plots confirmed that both BAIO and C₃N₄ were n-type semiconductors. For n-type semiconductors, the E_fb value was typically 0.2 V below the conduction band minimum (CBM).⁴⁴ Based on this, the CBM were calculated to be −1.00 V and −0.74 V (vs. NHE, pH = 0) for BAIO and C₃N₄, respectively. Combined with their E_g, the valence band maxima (VBM) for BAIO and C₃N₄ were determined to be 1.07 V and 2.27 V (vs. NHE, pH = 0), respectively. The energy difference between the Fermi level (E_f) and the VBM was further investigated by VB XPS.^36,37 As illustrated in Fig. 4e, the energy difference between the E_f and VBM of BAIO was 1.32 eV, while for C₃N₄ it was 2.00 eV (vs. NHE, pH = 0). Considering their VBM positions, the estimated E_f values for BAIO and C₃N₄ were −0.25 eV and 0.27 eV (vs. NHE, pH = 0), respectively. Based on the above analysis, the energy band diagram shown in Fig. 4f was obtained. The energy band structures of BAIO and C₃N₄ were staggered, with the E_f value of BAIO being higher than that of C₃N₄, which was consistent with the requirements for a Z-scheme heterojunction. The CB potentials of BAIO and C₃N₄ were more negative than the reduction potentials of O₂/H₂O₂ (+0.68 V vs. NHE, pH = 0) and O₂/˙O₂⁻ (+0.33 V vs. NHE, pH = 0), indicating that the reduction potentials were suitable for H₂O₂ production by the ORR. Moreover, the VB potential of C₃N₄ was more positive than the oxidation potential of H₂O/H₂O₂ (+1.76 V vs. NHE, pH = 0). This shows that the photocatalyst can also generate H₂O₂ through the WOR pathway.

Based on the aforementioned study, the Z-scheme charge transfer mechanism of the BAIO/C₃N₄ heterojunction is illustrated in Fig. 4g. When BAIO and C₃N₄ came into contact, the difference in their work functions caused e⁻ in BAIO to spontaneously transfer across the interface to C₃N₄ until the Fermi levels were aligned. It is worth mentioning that the direction of e⁻ transfer matched the results shown in the XPS spectrum in Fig. 2. As a result, BAIO became positively charged and C₃N₄ became negatively charged at the interface, which created an internal electric field (IEF) that pointed from BAIO to C₃N₄. Meanwhile, the energy band edges of BAIO and C₃N₄ bent in opposite directions. Upon illumination, the e⁻ in both BAIO and C₃N₄ was excited from the VB to the CB. The combined effects of the IEF, coulombic interactions, and energy band bending accelerated the transfer of the excited e⁻ in the CB of C₃N₄ toward the interface, where they recombined with the photogenerated holes (h⁺) in the VB of BAIO. Simultaneously, this process inhibited the recombination of e⁻ in the BAIO CB with h⁺ in the C₃N₄ VB. As a Z-scheme heterojunction, BAIO/C₃N₄ maintained the strong reducing ability of e⁻ in the CB of BAIO, which acted as the active site for the ORR.

As previously discussed, Z-scheme heterojunctions are known to generate a strong IEF, which significantly enhances charge separation and directional migration.^45–47 To validate this, a comparative analysis of the relative IEF strengths was performed using surface voltages measured by OCP measurements and surface charge densities derived from transient photocurrent response measurements. The BAIO/C₃N₄ system exhibited the highest OCP difference before and after light irradiation, reaching approximately 116 mV, which was substantially higher than those of BAIO (10 mV) and C₃N₄ (37 mV) (Fig. 5a). This suggested that BAIO/C₃N₄ achieved superior photogenerated charge separation efficiency.⁴⁸ By integrating the measured transient photocurrent density minus the steady-state photocurrent values with respect to time, the obtained value was proportional to the number of negative charges accumulated on the surface (Fig. S9†). The calculated surface charge densities for BAIO, C₃N₄ and BAIO/C₃N₄ were 0.089, 0.040, and 0.086 μC cm⁻², respectively. Using the model reported previously,⁴⁷ the IEF intensity in BAIO/C₃N₄ was found to be 3.4 times and 2.6 times higher than those in BAIO and C₃N₄, respectively (Fig. 5b).

The enhanced IEF promoted the separation and migration of charge carriers in the Z-scheme heterojunction. Fig. 5c presents the photoluminescence (PL) spectra of C₃N₄ and BAIO/C₃N₄, both of which showed strong emission in the range of 400 to 600 nm. Generally, a stronger PL intensity was associated with a higher charge recombination rate. Therefore, the lower peak intensity of BAIO/C₃N₄ implied the suppressed recombination of photogenerated charge carriers. In addition, BAIO/C₃N₄ exhibited a longer decay lifetime (τ) of 10.57 ns in time-resolved photoluminescence (TR-PL) measurements, compared to 8.36 ns for C₃N₄. This further indicated that the rapid carrier recombination was effectively suppressed in BAIO/C₃N₄ (Fig. 5d).⁴⁹ Nyquist plots of the EIS were subsequently measured to analyze the charge transfer properties of the prepared photocatalysts. As shown in Fig. 5e, BAIO/C₃N₄ revealed the smallest semicircular arc radius, indicating that the charge transfer resistance was reduced after the formation of the heterojunction. Thanks to the improved charge separation and migration processes, the BAIO/C₃N₄ heterojunction displayed the highest photocurrent density during the four intermittent light-switching cycles (Fig. 5f). The above results confirmed that the enhanced IEF effectively improved the separation and migration of the charge carrier in the BAIO/C₃N₄ heterojunction, which was one of the key factors in enhancing photocatalytic performance.

3.4. H₂O₂ generation mechanism

To clarify the mechanism of photocatalytic H₂O₂ generation, performance tests were initially carried out under different atmospheres (Fig. 6a). Compared to the O₂ atmosphere, the H₂O₂ yield measured in air showed a slight decrease. When the atmosphere was switched to N₂, only trace amounts of H₂O₂ were detected. These findings not only confirmed that O₂ was an essential reactant, but also proved that the ORR was indispensable in the overall process. Subsequently, active species trapping experiments for the BAIO/C₃N₄ heterojunction were conducted. Silver nitrate (AgNO₃), disodium ethylenediaminetetraacetic acid (EDTA-2Na), benzoquinone (BQ), and tert-butanol (TBA) were introduced to investigate the roles of e⁻, h⁺, ˙O₂⁻, and ˙OH in photocatalytic H₂O₂ production, respectively.³² As shown in Fig. 6b, in the presence of AgNO₃, the H₂O₂ yield of BAIO/C₃N₄ significantly decreased, indicating that the formation of H₂O₂ was primarily driven by the reduction of O₂ via e⁻. In contrast, the H₂O₂ yield increased after the addition of EDTA-2Na, which resulted from the accelerated electron-mediated ORR by the quenching of h⁺. The ˙O₂⁻ didn't participate in H₂O₂ production, as the H₂O₂ yield of BAIO/C₃N₄ showed little change when BQ was added. Thus, the ORR process might not follow an indirect two-step 2e⁻ ORR involving ˙O₂⁻. Since the addition of TBA had little effect on the activity, ˙OH also didn't take part in H₂O₂ production. Because the VB of C₃N₄ was not positive enough, it cannot generate ˙OH (OH⁻/˙OH, +2.73 V vs. NHE, pH = 0). As shown in Fig. S10,† the contact angles of C₃N₄ and BAIO/C₃N₄ were similar, suggesting that the performance was unaffected by hydrophilicity. These results proved that the generation of H₂O₂ on BAIO/C₃N₄ occurred via the ORR pathway.

The RRDE test was performed to determine the selectivity of the ORR and the number of electrons transferred (n) in the H₂O₂ production reaction.^12,50 The disk current that came from the ORR for the photocatalyst progressively increased as the applied potential decreased (Fig. 6c). At the same time, H₂O₂ generated on the disk electrode could rapidly diffuse into the ring electrode and be further oxidized to form a positive ring current. In addition, the n value for the reduction of O₂ to H₂O₂ was calculated. All three samples exhibited values close to the theoretical electron transfer number of 2, indicating their excellent kinetic selectivity for the formation of H₂O₂ from O₂ (Fig. 6d). The aforementioned active species trapping experiments demonstrated that ˙O₂⁻ didn't participate in the ORR pathway. To further verify this, electron paramagnetic resonance (EPR) experiments were conducted, using 5,5-dimethyl-1-pyrroline-N-oxide (DMPO) as the trapping agent. As shown in Fig. 6e, C₃N₄ displayed a distinct DMPO-˙O₂⁻ signal, while the DMPO-˙O₂⁻ signal was barely detectable in BAIO and BAIO/C₃N₄. These results collectively confirmed that H₂O₂ production in the BAIO/C₃N₄ heterojunction occurred via a direct one-step 2e⁻ ORR pathway, while C₃N₄ followed an indirect two-step 2e⁻ ORR process. To explore the reasons behind the change in the reaction pathway, in situ FTIR was employed to monitor the real-time intermediate products during the photocatalytic H₂O₂ production process (Fig. 6f and Fig. S11†). In Fig. 6f, the vibration of the –O–O– peak in the 1250–1100 cm⁻¹ range confirmed that the adsorption configuration of O₂ on BAIO/C₃N₄ was the Yeager-type.^13,51 Under light irradiation, the intensity of Yeager-type O₂ increased as the photocatalytic reaction progressed. The peak corresponding to H₂O₂ at 1280 cm⁻¹ also intensified during the photocatalytic reaction, demonstrating that H₂O₂ was produced by the photocatalysis. In contrast, the Yeager-type O₂ adsorption was not observed in the C₃N₄ spectra (Fig. S11a†). Additionally, the signals for Yeager-type O₂ and H₂O₂ were not detected, likely due to the relatively low activity of BAIO (Fig. S11b†). Overall, the incorporation of BAIO led to the Yeager-type O₂ adsorption at the BAIO/C₃N₄ surface, which effectively facilitated the photocatalytic conversion of O₂ to H₂O₂ via a direct one-step 2e⁻ ORR.^13,52

Based on the above experimental results, the enhanced photocatalytic H₂O₂ production performance of the BAIO/C₃N₄ heterojunction could be explained. The Z-scheme heterojunction formed between BAIO and C₃N₄ facilitated charge separation and migration through an enhanced IEF. Meanwhile, the O₂ adsorption configuration on the surface was altered to the Yeager-type, which not only prevented the formation of highly oxidizing ˙O₂⁻ intermediates but also enabled the direct generation of H₂O₂ from O₂ via a direct one-step 2e⁻ ORR. Collectively, these improvements significantly enhanced the overall photocatalytic efficiency for H₂O₂ production.

4. Conclusion

In summary, the Z-scheme BAIO/C₃N₄ heterojunction was successfully constructed using an in situ synthesis strategy. The formation of the heterojunction improved the charge separation and migration facilitated by an enhanced IEF. Additionally, it altered the O₂ adsorption configuration on the photocatalyst to the Yeager-type, thereby facilitating the transition of the reaction pathway from the indirect two-step 2e⁻ ORR to the direct one-step 2e⁻ ORR. Thanks to these advantages, the BAIO/C₃N₄ heterojunction with a BAIO:C₃N₄ ratio of 1:1.5 achieved a H₂O₂ yield of 535.9 μmol g⁻¹ h⁻¹ under visible light irradiation, representing an 8.8-fold enhancement in performance compared to pure C₃N₄. This study reveals the importance of the direct one-step 2e⁻ ORR for enhancing photocatalytic H₂O₂ production activity and offers a valuable approach to regulate the reaction pathway.

Author contributions

K. Y. and Z. S. contributed equally to this work. Q. K., H. X. L. and Z. X. X. initialized this research. K. Y. and Z. S. designed and discussed the experiments. K. Y., Z. S., T. Z. and Y. H. analyzed the data. K. Y., W. Y. and Z. S. conducted and interpreted the original infrared test. K. Y., Z. S. and X. F. drafted the manuscript. Q. K. and H. X. L. supervised the research and revised the manuscript.

Data availability

The data supporting this article have been included as part of the ESI.†

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (22275153, 22371236, 22275155, 21931009, and U24A20563), the Natural Science Foundation of Xiamen, China (3502Z20227008), and the Fundamental Research Funds for the Central Universities (20720220011).

References

1. Y. Y. Sun, L. Han and P. Strasser, A comparative perspective of electrochemical and photochemical approaches for catalytic H₂O₂ production, Chem. Soc. Rev., 2020, 49, 6605–6631 RSC .
2. H. L. Hou, X. K. Zeng and X. W. Zhang, Production of hydrogen peroxide by photocatalytic processes, Angew. Chem., Int. Ed., 2020, 59, 17356–17376 CrossRef CAS PubMed .
3. M. Song, W. Liu, J. J. Zhang, C. Zhang, X. Huang and D. L. Wang, Single-atom catalysts for H₂O₂ electrosynthesis via two-electron oxygen reduction reaction, Adv. Funct. Mater., 2023, 33, 2212087 CrossRef CAS .
4. Z. Chen, D. C. Yao, C. C. Chu and S. Mao, Photocatalytic H₂O₂ production systems: Design strategies and environmental applications, Chem. Eng. J., 2023, 451, 138489 CrossRef CAS .
5. Z. J. Yong and T. Y. Ma, Solar-to-H₂O₂ catalyzed by covalent organic frameworks, Angew. Chem., Int. Ed., 2023, 62, e202308980 CrossRef CAS PubMed .
6. X. Zhang, H. Su, P. X. Cui, Y. Y. Cao, Z. Y. Teng, Q. T. Zhang, Y. Wang, Y. B. Feng, R. Feng, J. X. Hou, X. Y. Zhou, P. J. Ma, H. W. Hu, K. W. Wang, C. Wang, L. Y. Gan, Y. X. Zhao, Q. H. Liu, T. R. Zhang and K. Zheng, Developing Ni single-atom sites in carbon nitride for efficient photocatalytic H₂O₂ production, Nat. Commun., 2023, 14, 7115 CrossRef CAS PubMed .
7. H. Yu, F. T. Zhang, Q. Chen, P.-K. Zhou, W. D. Xing, S. B. Wang, G. G. Zhang, Y. Jiang and X. Chen, Vinyl-group-anchored covalent organic framework for promoting the photocatalytic generation of hydrogen peroxide, Angew. Chem., Int. Ed., 2024, 63, e202402297 CrossRef CAS PubMed .
8. Q. Zhang, H. N. Che, H. B. Yang, B. Liu and Y. H. Ao, Tandem proton transfer in carboxylated supramolecular polymer for highly efficient overall photosynthesis of hydrogen peroxide, Angew. Chem., Int. Ed., 2024, 63, e202409328 CrossRef CAS PubMed .
9. S. Wu and X. Quan, Design principles and strategies of photocatalytic H₂O₂ production from O₂ reduction, ACS ES&T Eng., 2022, 2, 1068–1079 Search PubMed .
10. L. J. Liu, M.-Y. Gao, H. F. Yang, X. Y. Wang, X. B. Li and A. I. Cooper, Linear conjugated polymers for solar-driven hydrogen peroxide production: The importance of catalyst stability, J. Am. Chem. Soc., 2021, 143, 19287–19293 CrossRef CAS PubMed .
11. D. X. Jiao, C. S. Ding, M. H. Xu, X. W. Ruan, S. K. Ravi and X. Q. Cui, Modulating yeager adsorption configuration of O₂ through Cd doping in Zn₃In₂S₆ for photosynthesis of H₂O₂, Adv. Funct. Mater., 2024, 35, 2416753 CrossRef .
12. P. Liu, T. Liang, Y. T. Li, Z. Q. Zhang, Z. Li, J. Bian and L. Q. Jing, Photocatalytic H₂O₂ production over boron-doped g-C₃N₄ containing coordinatively unsaturated FeOOH sites and CoO_x clusters, Nat. Commun., 2024, 15, 9224 CrossRef CAS PubMed .
13. Y. Luo, B. P. Zhang, C. C. Liu, D. H. Xia, X. W. Ou, Y. P. Cai, Y. Zhou, J. Jiang and B. Han, Sulfone-modified covalent organic frameworks enabling efficient photocatalytic hydrogen peroxide generation via one-step two-electron O₂ reduction, Angew. Chem., Int. Ed., 2023, 62, e202305355 CrossRef CAS PubMed .
14. X. H. Huang, C. Q. Zhou, J. H. Yang, P. P. Sun, Y. H. Song, R. Y. Ma, X. X. Ding and Z. Mo, Potassium/cyano group co-incorporation promotes 2e⁻ ORR selectivity in porous ultrathin carbon nitride for photocatalytic H₂O₂ production, Solid State Sci., 2024, 156, 107669 CrossRef CAS .
15. Y. Shiraishi, S. Kanazawa, Y. Sugano, D. Tsukamoto, H. Sakamoto, S. Ichikawa and T. Hirai, Highly selective production of hydrogen peroxide on graphitic carbon nitride (g-C₃N₄) photocatalyst activated by visible light, ACS Catal., 2014, 4, 774–780 CrossRef CAS .
16. L. X. Xie, X. Y. Wang, Z. Y. Zhang, Y. Y. Ma, T. Du, R. Wang and J. L. Wang, Photosynthesis of hydrogen peroxide based on g-C₃N_4: The road of a cost-effective clean fuel production, Small, 2023, 19, 25 Search PubMed .
17. T. Y. Gao, D. G. Zhao, S. S. Yuan, M. Zheng, X. J. Pu, L. Tang and Z. D. Lei, Energy band engineering of graphitic carbon nitride for photocatalytic hydrogen peroxide production, Carbon Energy, 2024, 6, e596 CrossRef CAS .
18. S. K. Sahoo, L. Acharya, L. Biswal, P. Priyadarshini and K. Parida, Recent advancements in graphitic carbon nitride based direct Z- and S-scheme heterostructures for photocatalytic H₂O₂ production, Inorg. Chem. Front., 2024, 11, 4914–4973 RSC .
19. Y.-X. Ye, J. H. Pan, F. Y. Xie, L. Gong, S. M. Huang, Z. F. Ke, F. Zhu, J. Q. Xu and G. F. Ouyang, Highly efficient photosynthesis of hydrogen peroxide in ambient conditions, Proc. Natl. Acad. Sci. U. S. A., 2021, 118, e2103964118 CrossRef CAS PubMed .
20. L. Chen, C. Chen, Z. Yang, S. Li, C. C. Chu and B. L. Chen, Simultaneously tuning band structure and oxygen reduction pathway toward high-efficient photocatalytic hydrogen peroxide production using cyano-rich graphitic carbon nitride, Adv. Funct. Mater., 2021, 31, 2105731 CrossRef CAS .
21. Q. C. Chen, C. J. Lu, B. Y. Ping, G. Y. Li, J. Y. Chen, Z. M. Sun, Y. J. Zhang, Q. S. Ruan and L. Tao, A hydroxyl-induced carbon nitride homojunction with functional surface for efficient photocatalytic production of H₂O₂, Appl. Catal., B, 2023, 324, 122216 CrossRef CAS .
22. Y. Z. Qian, J. Q. Wu, Y. W. Zhang, Z. J. Wang, W. K. Bao, X. F. Zeng and Q. H. Zhu, Recyclable hydrogen peroxide photosynthesis over Z-scheme In₂O₃/g-C₃N₄ heterojunction with viable reduction of superoxide radical intermediate, Appl. Surf. Sci., 2025, 680, 161302 CrossRef CAS .
23. J. Y. An, W. J. Jiang, F. W. Zhuang, Y. H. Ma, S. Zhan and F. Zhou, Boosting solar-to-H₂O₂ by molecularly tunable heterostructured Pym-CN with an enhanced built-in electric field, ACS Catal., 2024, 14, 17014–17024 CrossRef CAS .
24. Q. Zhang, K. Y. Gu, C. R. Dong, C. Xue, H. N. Che, K. Zhang and Y. H. Ao, Polymeric carbon nitride edged with spatially isolated donor and acceptor for sunlight-driven H₂O₂ synthesis and in situ utilization, Angew. Chem., Int. Ed., 2024, 64, e202417591 CrossRef PubMed .
25. J. Y. Qiu, K. Meng, Y. Zhang, B. Cheng, J. J. Zhang, L. X. Wang and J. G. Yu, COF/In₂S₃ S-scheme photocatalyst with enhanced light absorption and H₂O₂-production activity and fs-TA investigation, Adv. Mater., 2024, 36, 2400288 CrossRef CAS PubMed .
26. W. Y. Yu, Z. J. Zhu, C. Hu, S. Lin, Y. H. Wang, C. Y. Wang, N. Tian, Y. H. Zhang and H. W. Huang, Point-to-faceZ-scheme junction Cd_0.6Zn_0.4S/g-C₃N₄with a robust internal electric field for high-efficiency H₂O₂production, J. Mater. Chem. A, 2023, 11, 6384–6393 RSC .
27. H. C. Ma, Y. M. Wang, Z. A. Zhang, J. J. Liu, Y. C. Yu, S. L. Zuo and B. S. Li, A superior ternary Z-scheme photocatalyst of Bi/Black phosphorus nanosheets/P-doped BiOCl containing interfacial P-P bond and metallic mediator for H₂O₂ production and RhB degradation, Chemosphere, 2023, 330, 138717 CrossRef CAS PubMed .
28. Z. Y. Liao, J. J. Du, L. Wang, M. Yang, L. L. Zhao, X. Q. Zhan, H. L. Yang, W. Y. Yang and H. L. Hou, Advances in Z-scheme photocatalysts for efficient photocatalytic H₂O₂ production: A comprehensive review, J. Alloys Compd., 2024, 976, 173322 CrossRef CAS .
29. G. F. Liao, C. X. Li, X. Z. Li and B. Z. Fang, Emerging polymeric carbon nitride Z-scheme systems for photocatalysis, Cell Rep. Phys. Sci., 2021, 2, 100355 CrossRef CAS .
30. G. Volonakis, N. Sakai, H. J. Snaith and F. Giustino, Oxide analogs of halide perovskites and the new semiconductor Ba₂AgIO₆, J. Phys. Chem. Lett., 2019, 10, 1722–1728 CrossRef CAS PubMed .
31. S. Bazrafshan, J. He and S. S. Naghavi, Charting Ba-based double perovskite oxides for visible-light-driven photocatalytic water splitting, J. Phys. Chem. C, 2023, 127, 3968–3976 CrossRef CAS .
32. X. Zhang, P. J. Ma, C. Wang, L. Y. Gan, X. J. Chen, P. Zhang, Y. Wang, H. Li, L. H. Wang, X. Y. Zhou and K. Zheng, Unraveling the dual defect sites in graphite carbon nitride for ultra-high photocatalytic H₂O₂ evolution, Energy Environ. Sci., 2022, 15, 830–842 RSC .
33. X. J. Chen, R. Shi, Q. Chen, Z. J. Zhang, W. J. Jiang, Y. F. Zhu and T. R. Zhang, Three-dimensional porous g-C₃N₄ for highly efficient photocatalytic overall water splitting, Nano Energy, 2019, 59, 644–650 CrossRef CAS .
34. E. A. Pogue, J. Bond, C. Imperato, J. B. S. Abraham, N. Drichko and T. M. McQueen, A gold(I) oxide double perovskite: Ba₂AuIO₆, J. Am. Chem. Soc., 2021, 143, 19033–19042 CrossRef CAS PubMed .
35. C. K. Fan, J. L. Miao, G. Q. Xu, J. Q. Liu, J. Lv and Y. C. Wu, Graphitic carbon nitride nanosheets obtained by liquid stripping as efficient photocatalysts under visible light, RSC Adv., 2017, 7, 37185–37193 RSC .
36. Q. Chen, Y. Zhang, E. M. You, Q. R. Jiang, X. J. Chen, Y. Wang, Z. J. Song, K. Chang, Z. X. Xie and Q. Kuang, Accelerated water oxidation kinetics triggered by supramolecular porphyrin nanosheet for robust visible-light-driven CO₂ reduction, Small, 2022, 18, 2204924 CrossRef CAS PubMed .
37. Z. J. Song, Q. Chen, Z. W. Sun, K. Chang, Z. X. Xie and Q. Kuang, Constructing Z-scheme WO₃/C₃N₄ heterojunctions with an enlarged internal electric field and accelerated water oxidation kinetics for robust CO₂ photoreduction, J. Mater. Chem. A, 2024, 12, 14426–14436 RSC .
38. J. W. Fu, Q. L. Xu, J. X. Low, C. J. Jiang and J. G. Yu, Ultrathin 2D/2D WO₃/g-C₃N₄ step-scheme H₂-production photocatalyst, Appl. Catal., B, 2019, 243, 556–565 CrossRef CAS .
39. D. K. Smith, J. McCollum and M. L. Pantoya, Effect of environment on iodine oxidation state and reactivity with aluminum, Phys. Chem. Chem. Phys., 2016, 18, 11243–11250 RSC .
40. Y. X. Yang, L. Geng, Y. N. Guo, J. Q. Meng and Y. H. Guo, Easy dispersion and excellent visible-light photocatalytic activity of the ultrathin urea-derived g-C₃N₄ nanosheets, Appl. Surf. Sci., 2017, 425, 535–546 CrossRef CAS .
41. T. Y. Liu, F. Chen, Y. An, H. W. Huang, J. G. Liu, W. Y. Yu, M. T. Li, L. Q. Bai, Y. H. Zhang and N. Tian, Oxygen vacancy induced robust interfacial electric field for efficient photocatalytic hydrogen peroxide production, Chem. Eng. J., 2024, 479, 147724 CrossRef CAS .
42. Z. B. Fang, T. T. Liu, J. Liu, S. Jin, X. P. Wu, X. Q. Gong, K. Wang, Q. Yin, T. F. Liu, R. Cao and H. C. Zhou, Boosting interfacial charge-transfer kinetics for efficient overall CO₂ photoreduction via rational design of coordination spheres on metal-organic frameworks, J. Am. Chem. Soc., 2020, 142, 12515 CrossRef CAS PubMed .
43. G. G. Liu, T. Wang, S. X. Ouyang, L. Q. Liu, H. Y. Jiang, Q. Yu, T. Kako and J. H. Ye, Band-structure-controlled BiO(ClBr)_(1−x)/2I_x solid solutions for visible-light photocatalysis, J. Mater. Chem. A, 2015, 3, 8123–8132 RSC .
44. X. W. Shi, L. A. Mao, P. Yang, H. J. Zheng, M. Fujitsuka, J. Y. Zhang and T. Majima, Ultrathin ZnIn₂S₄ nanosheets with active (110) facet exposure and efficient charge separation for cocatalyst free photocatalytic hydrogen evolution, Appl. Catal., B, 2020, 265, 118616 CrossRef CAS .
45. X. Y. Yue, J. J. Fan and Q. J. Xiang, Internal electric field on steering charge migration: Modulations, determinations and energy-related applications, Adv. Funct. Mater., 2021, 32, 2110258 CrossRef .
46. Y. Guo, W. X. Shi and Y. F. Zhu, Internal electric field engineering for steering photogenerated charge separation and enhancing photoactivity, EcoMat, 2019, 1, e12007 CrossRef .
47. Q. Chen, X. J. Chen, Q. R. Jiang, Z. P. Zheng, Z. J. Song, Z. Y. Zhao, Z. X. Xie and Q. Kuang, Constructing oxide/sulfide in-plane heterojunctions with enlarged internal electric field for efficient CO₂ photoreduction, Appl. Catal., B, 2021, 297, 120394 CrossRef CAS .
48. Y. X. Shi, L. L. Li, Z. Xu, X. L. Qin, Y. Cai, W. L. Zhang, W. L. Shi, X. Du and F. Guo, Coupled internal electric field with hydrogen release kinetics for promoted photocatalytic hydrogen production through employing carbon coated transition metal as co-catalyst, J. Colloid Interface Sci., 2023, 630, 274–285 CrossRef CAS PubMed .
49. X. S. Zhao, Y. Y. You, S. B. Huang, Y. X. Wu, Y. Y. Ma, G. Zhang and Z. H. Zhang, Z-scheme photocatalytic production of hydrogen peroxide over Bi₄O₅Br₂/g-C₃N₄ heterostructure under visible light, Appl. Catal., B, 2020, 278, 119251 CrossRef CAS .
50. R. F. Zhou, Y. Zheng, M. Jaroniec and S. Z. Qiao, Determination of the electron transfer number for the oxygen reduction reaction: From theory to experiment, ACS Catal., 2016, 6, 4720–4728 CrossRef CAS .
51. H. Y. Chen, H. Zhang, K. Chi and Y. Zhao, Pyrimidine-containing covalent organic frameworks for efficient photosynthesis of hydrogen peroxide via one-step two electron oxygen reduction process, Nano Res., 2024, 17, 9498–9506 CrossRef CAS .
52. W. T. Li, B. Han, Y. H. Liu, J. Y. Xu, H. R. He, G. G. Wang, J. S. Li, Y. X. Zhai, X. L. Zhu and Y. F. Zhu, Unsymmetric protonation driven highly efficient H₂O₂ photosynthesis in supramolecular photocatalysts via one-step two-electron oxygen reduction, Angew. Chem., Int. Ed., 2024, 64, e202421356 CrossRef PubMed .

---

Footnotes

† Electronic supplementary information (ESI) available: Experimental results including SEM images, TEM images, XRD patterns, XPS spectra, LSV curves, and N2 adsorption–desorption isotherms; water contact angle tests and transient photocurrent density tests of the materials; and in situ DRIFTS spectra. See DOI: https://doi.org/10.1039/d5qi01052g

‡ These authors contributed equally to this work.

This journal is © the Partner Organisations 2025