Layered bismuth oxyhalides (commonly denoted as BiOX, where X = Cl, Br, I) are typical representatives of this structural family. Their crystal structure consists of alternating cationic Bi₂O₂²⁺ layers and anionic halogen ion (X^-) slabs^[71], which are stacked through weak van der Waals interactions, as shown in Figure 3A. This structural configuration yields a relatively high specific surface area along with abundant exposed active sites. Meanwhile, the band gap progressively narrows as the atomic number of the halogen increases (Eg, BiOCl > BiOBr > BiOI), thereby enhancing visible-light absorption. Notably, the weak interlayer bonding allows facile modification of bismuth oxyhalides through strategies such as ion exchange or intercalation (e.g., incorporation of organic molecules or metal nanoparticles). Moreover, an intrinsic static electric field is established between the positively charged Bi₂O₂²⁺ layers and the negatively charged X^- layers^[39], which facilitates the separation of photogenerated electrons (e^-) and holes (h⁺) while suppressing their recombination. Charge transport exhibits strong anisotropy: electrons migrate predominantly within the layers (perpendicular to the c-axis), whereas holes transport across the interlayer space, thereby improving charge carrier utilization efficiency. The layered architecture of bismuth oxyhalides offers versatile routes for material engineering, including heterojunction construction, surface defect incorporation, and morphological nanostructuring^[72,73].

Bi₂MoO₆ and Bi₂WO₆ are layered Aurivillius-type oxides, consisting of alternating stacks of positively charged Bi₂O₂²⁺ layers and perovskite-like MO₆^6- slabs (M = Mo, W), as shown in Figure 3B. The 6s² lone-pair electrons of Bi³⁺ ions, being stereochemically active, give rise to spontaneous polarization, generating a strong internal electric field between the layers. This built-in field effectively promotes the unidirectional separation of photogenerated charge carriers, with electrons moving toward the perovskite-like slabs and holes toward the Bi₂O₂²⁺ layers, thereby significantly suppressing recombination^[74]. These materials exhibit a moderate band gap (approximately 2.4-2.9 eV), providing good visible-light absorption. The weak interlayer interactions facilitate interlayer charge transfer, enabling fast movement of photogenerated electrons from the perovskite-like slabs toward surface N₂ adsorption sites and reducing carrier recombination. The layered architecture provides abundant surface-active sites, and hybridization between Bi 6s and N 2p orbitals improves the kinetics of nitrogen chemisorption and activation. Specifically, in Bi₂MoO₆, significant distortion of the MoO₆ octahedra synergizes with the lone pair effect of Bi³⁺, reinforcing the internal electric field^[42]. In Bi₂WO₆, the distortion of the WO₆ octahedra is less pronounced; however, its ultrathin nanosheet morphology shortens charge migration pathways, further suppressing bulk recombination and exposing more active facets. Although interlayer coupling is relatively strong, strategies such as constructing heterojunctions can effectively tailor their catalytic performance^[75].

Perovskites have the general formula ABO₃, as shown in Figure 3C. In bismuth-based catalysts, bismuth (Bi) can occupy different sites (e.g., Bi_0.5Na_0.5TiO₃ and SrBiO₃). In either position, its inclusion fundamentally alters the material’s intrinsic properties. The bismuth ion (Bi³⁺) has a stereochemically active 6s² lone pair of electrons. At the A-site, this lone pair induces pronounced local lattice distortion, breaking centrosymmetry. The distortion can generate internal electric fields and destabilize lattice oxygen, promoting oxygen vacancy formation. When bismuth occupies the B-site in higher valence states (e.g., Bi⁴⁺/Bi⁵⁺), it acts as a direct catalytic center. The Bi³⁺/Bi⁵⁺ redox couple provides efficient electron-transfer pathways for reactions. Notably, some bismuth-based perovskites (e.g., BiFeO₃^[76,77]) may undergo surface corrosion under mild conditions, leading to bismuth dissolution and the in situ formation of a highly active (hydr)oxide layer. This self-reconstruction creates a core-shell architecture: a stable perovskite core serves as a conductive skeleton, while the surface (hydr) oxide provides abundant active sites.

Pyrochlore-structured materials, exemplified by Bi₂X₂O₇ (X = Sn, Ti, Zr), feature a three-dimensional network composed of alternating Bi₂O₂²⁺ layers and oxygen-based anionic groups, as shown in Figure 3D. In this structure, Bi³⁺ ions adopt an eight-coordinate geometry, while X⁴⁺ ions occupy distinct crystallographic sites^[78-81]. The good match in ionic radii between the cations, combined with their linkage via oxygen bridges, confers excellent thermal and chemical stability to the framework. These materials exhibit a moderate band gap (typically 2.0-3.0 eV), fulfilling essential criteria for photocatalytic nitrogen fixation. Cation vacancies and a continuous oxygen-coordination network serve as efficient pathways for the transport of photogenerated electrons. Oxygen vacancies, as a prominent type of surface defect, function as active sites for nitrogen adsorption and activation. They promote N≡N bond cleavage by destabilizing the triple bond, thereby reducing the kinetic barrier for subsequent reduction^[82]. Furthermore, pyrochlore-type compounds are often coupled with semiconductors such as TiO₂ and g-C₃N₄ to form heterojunctions^[83,84]. This approach extends the light-harvesting range and leverages the interfacial internal electric field to promote charge separation, leading to enhanced.

Overall photocatalytic efficiency. The structure also exhibits considerable tolerance to doping and defects, allowing its electronic structure and catalytic activity to be systematically tuned through compositional design.

In a Type II heterojunction, the energy bands of the two semiconductors are staggered, creating a mismatched alignment. Photogenerated electrons tend to move from the semiconductor with a higher CBM to the one with a lower CBM. Conversely, photogenerated holes migrate from the semiconductor with a lower valence band maximum (VBM) to the one with a higher VBM. This spatial separation of charges effectively reduces bulk recombination of electron-hole pairs and markedly extends carrier lifetimes - an important benefit of Type II heterojunctions in photocatalytic applications.

However, this charge separation incurs a reduction in some redox potential. Specifically, electrons localize on the semiconductor with a higher CBM (indicating a weaker reduction ability), while holes accumulate on the semiconductor with a lower VBM (indicating a weaker oxidation ability). Consequently, electrons and holes tend to become confined at sites where their reducing and oxidizing capabilities are comparatively limited. This results in a decreased driving force per charge carrier, despite an increased number of carriers available for surface reactions. Therefore, designing a Type II heterojunction requires balancing charge-separation efficiency with maintaining adequate redox potentials. When the band alignment of the two semiconductors is properly matched - even if electrons transfer to a slightly less favorable reduction site - the conduction band minimum can still be positioned above the thermodynamic potential needed for specific reactions such as N₂ reduction to NH₃^[37]. In this case, photocatalytic efficiency is greatly enhanced while ensuring the reaction proceeds.

Constructing a Type-II heterojunction requires precise calculation and design of the band positions of the two semiconductors to achieve an ideal staggered alignment. Moreover, it is essential to form intimate, large-area heterointerfaces, such as core-shell structures or layered composites, to provide ample channels for the directional migration of electrons and holes. The synthesis process must be finely controlled to minimize interfacial defects, since the core of interface engineering in Type-II heterojunctions lies in ensuring good physical contact and proper band alignment while suppressing interfacial defects^[101]. Table 1 summarizes data on type-II heterojunctions over the past few years.

BiPO₄ is a low-cost, non-toxic, and chemically stable Ultraviolet-Visible (UV)-responsive photocatalyst with activity surpassing that of TiO₂. It has been widely employed as a cocatalyst to promote charge separation in a heterostructure. Zhao et al.^[111] synthesized a BiPO₄/Bi₄O₅Br₂ composite photocatalyst via a simple solvothermal method, constructing a Type II heterojunction at the interface. Due to the higher Fermi level of BiPO₄ compared to Bi₄O₅Br₂, electron transfer from BiPO₄ to Bi₄O₅Br₂ occurs upon contact, resulting in negative charge accumulation in the interfacial region of Bi₄O₅Br₂ and positive charge in BiPO₄. The charge redistribution establishes a built-in electric field that enables directional migration of photogenerated carriers, as shown in Figure 5A, thereby achieving efficient spatial separation of electron-hole pairs. As a key Bi-based semiconductor, Bi₄O₅Br₂ features a layered structure composed of alternating [Bi₂O₂]²⁺ and halogen-ion layers, which generates a built-in electric field along the c-axis that promotes charge separation upon photogeneration. Nonetheless, rapid carrier recombination in pure Bi₄O₅Br₂ limits the availability of electrons for nitrogen reduction, constraining its nitrogen fixation performance. To address this, Yue et al.^[109] Fabricated a KNbO₃/Bi₄O₅Br₂ heterojunction by embedding KNbO₃ microcubes into Bi₄O₅Br₂ nanosheets via a combined hydrothermal - precipitation approach, achieving intimate interfacial contact. The resulting charge redistribution induces downward band bending in Bi₄O₅Br₂ and an upwardly bent band structure in KNbO₃, which facilitates electron transfer from the CB of Bi₄O₅Br₂ to that of KNbO₃, and holes from the VB of KNbO₃ to that of Bi₄O₅Br₂, as shown in Figure 5B, thereby enabling efficient spatial separation of electron-hole pairs.

Other bismuth-based semiconductors, such as BiOBr, Bi₂WO₆, Bi₂MoO_6, and Bi(OH)₃, have also shown promise in photocatalysis. Among them, bismuth hydroxide [Bi(OH)₃] exhibits notable activity, but its wide bandgap (3.40 eV) restricts light absorption to the UV region. To address this, Liu et al.^[108] synthesized a BrO₃^--bridged Bi₂O₃/Bi(OH)₃ heterojunction via a hydrothermal method under strongly alkaline conditions. At the interface, BrO₃^- forms a Bi-BrO₃^--O-Bi structure that serves as an effective charge-transfer bridge. This configuration captures photogenerated electrons from the conduction band of Bi₂O₃, facilitating the reduction of Br and increased local electron density, as shown in Figure 5C. Consequently, charge transport resistance is reduced, and interfacial charge-transfer efficiency is significantly enhanced. Cao et al.^[112] constructed a ternary Bi₁₂O₁₇Cl₂/BiOCOOH/Bi₂MoO₆ heterojunction via a one-step synthesis under mild conditions, achieving both heterojunction formation and the introduction of oxygen vacancies. Bi₁₂O₁₇Cl₂ nanoparticles are uniformly deposited onto the precursor surface without altering its morphology, forming an intimate interfacial contact with no physical gaps. In term of charge transfer, electrons migrate sequentially from the CB of Bi₂MoO₆ to those of BiOCOOH and Bi₁₂O₁₇Cl₂ in a stepwise energy increase, while holes transfer in the reverse direction, moving from the VB of Bi₁₂O₁₇Cl₂ to BiOCOOH and Bi₂MoO₆, enabling effective spatial separation of charge carriers., enabling effectively spatial separation of charge carriers, as shown in Figure 5D.

Using an in-situ solvothermal approach, Xue et al.^[114] decorated Bi₂MoO₆ nanorods with oxygen vacancy-rich p-type BiOBr nanosheets (OV-BiOBr), forming a hierarchical “nanorod-nanosheet” heterointerface, as shown in Figure 6A-C. The close contact between crystal facets and the type-II band alignment achieved through p-n junction engineering facilitates the migration of photogenerated electrons from the CB of Bi₂MoO₆ to that of OV-BiOBr and the migration of holes from the valence band of OV-BiOBr to that of Bi₂MoO₆, significantly suppressing electron-hole recombination. Moreover, the introduced OVs serve as electron traps at the interface, capturing transferred electrons and further reducing charge recombination, as shown in Figure 6D. Zhao et al.^[105] synthesized Bi₄O₅Br₂/CdWO₄ composite via a two-step hydrothermal method, ensuring intimate interfacial contact rather than simple physical mixing. The resulting charge redistribution is accompanied by upward band bending in CdWO₄ and downward band bending in Bi₄O₅Br₂. This synergistic alignment greatly enhances the transfer of photogenerated electrons from the conduction band of Bi₄O₅Br₂ to that of CdWO₄ while promoting hole migration from the valence band of CdWO₄ to the valence band of Bi₄O₅Br₂, resulting in efficient spatial separation of charge carriers, as illustrated in Figure 6E. The type-II heterojunction also minimizes electron-hole recombination, which is supported by increased photocurrent, lowered charge-transfer resistance observed in electrochemical impedance spectroscopy, and notable suppression of photoluminescence emission, as depicted in Figure 6F-H.

Overall, Type-II heterojunctions offer a straightforward and effective interface engineering approach to spatially separate photogenerated carriers, although this may slightly weaken redox potentials. Their performance depends critically on close interfacial contact, staggered band alignment, and suppression of interfacial defects, which together improve the availability of electrons for nitrogen reduction.

In the Z-scheme heterojunction structure, photogenerated electrons (e^-) accumulate in the CB of one component and engage in reduction half-reactions. Since the conduction band position of this component is higher than that of the other, these electrons have a stronger reduction potential. As a result, bismuth-based Z-scheme heterojunctions generally demonstrate superior photocatalytic reduction performance compared to type-II heterojunctions, making them a primary focus in photocatalytic nitrogen fixation. Although less structurally complex than type-II heterojunctions, Z-scheme heterojunctions still necessitate close contact between the semiconductors and the intermediate conductor to facilitate effective electron transfer. The core strategy in interface engineering for Z-scheme heterojunctions is therefore to construct a conductive “bridge” that directs the selective recombination of charge carriers. This mechanism ensures the spatially directed recombination of less active charge carriers, thereby retaining holes with the highest oxidation potential in the component with a lower valence band position, and electrons with the highest reduction potential in the component with a higher conduction band position^[115]. Table 2 summarizes data on Z-type heterojunctions over the past few years.

Liu et al.^[56] directly anchored 2D boron-doped graphene quantum dots (BGQDs) onto the surface of 3D Bi₂MoO₆ (BMO) microspheres using an in-situ growth method, forming a closely integrated “quantum dot-microsphere” structure, as shown in Figure 7A-C. The BGQDs are bonded to surface atoms of BMO via chemical bonds, such as B-C and B-O, thereby avoiding interface detachment issues associated with physical mixing. The interfacial chemical bonding also suppresses the agglomeration tendency of BGQDs, as shown in Figure 7D-F. Even when the loading of BGQDs is increased to 30%, only a slight performance decline is observed, without severe light-blocking. Cao et al.^[116] utilized interlayer van der Waals forces to spontaneously form a 2D/2D stacked heterojunction of Bi₂MoO₆/g-C₃N₄ (BMO/CN). This process requires no additional binders or high-temperature treatment, thereby avoiding interface contamination and structural damage, and ensuring the cleanliness and integrity of the interface. Differential charge density calculations reveal significant charge redistribution at the CN-BMO interface, as shown in Figure 7G. Bader charge analysis indicates that approximately 0.92 e is transferred from CN to BMO, confirming the directionality of electron transfer at the interface. The bridging N atoms of CN at the interface serve as electron migration channels, facilitating efficient interfacial charge transport, as shown in Figure 7H-I.

Noble metals (e.g., Pt, Au, Ag) have been widely used as electron mediators in traditional Z‑scheme heterojunction photocatalysts^[12,127,128] but these are costly and hinder large-scale applications. Moreover, in photocatalytic nitrogen fixation, the HER strongly competes with the nitrogen reduction reaction (NRR), thereby reducing the efficiency of NH₃ production. To address these issues, Chen et al.^[118] Co-deposited Bi metal and Bi₂O₃ onto the surface of KTa_0.5Nb_0.5O₃ (KTN) via a one-step solvothermal method, forming a Bi-Bi₂O₃/KTN ternary composite system, rather than a modification with only Bi or Bi₂O₃. The incomplete core-shell structure plays a key interfacial role, as shown in Figure 8A and B. It ensures compatibility between KTN and Bi₂O₃, while the exposed Bi metal serves as active sites for N₂ adsorption, suppresses HER, and enhances NRR. Upon photoexcitation, electrons from the CB of Bi₂O₃are transferred through the Bi metal bridge to the VB of KTN, where they recombine with holes. Meanwhile, the electrons retained in the conduction band of KTN participate in N₂ reduction, as shown in Figure 8C. Through tailoring the composition and structure of the interface, the selectivity of interfacial reactions is improved. Owing to its high hydrogen adsorption energy and low HER activity, the exposed Bi metal suppresses the reduction of H⁺ to H₂ at the interface, thereby minimizing competitive H₂ evolution against NRR. Moreover, Bi metal can effectively adsorb and activate the N≡N triple bond. Through its synergistic interaction with KTN, the energy barrier for N₂ reduction is lowered.

Most photocatalysts, including TiO₂, g-C₃N₄^[129,130], and BiOX (X = Cl, Br, I), are constrained ed by rapid recombination of photogenerated electron-hole pairs, which severely limits the number of charge carriers available for N₂ reduction. Despite its suitable band structure and excellent chemical stability, which make Bi₄O₅Br₂ promising for photocatalytic nitrogen fixation, the single-component system still suffers from fast charge recombination. To overcome this limitation, Wang et al.^[117] constructed a direct Z-scheme heterojunction between BiOBr and Bi₄O₅Br₂ via a one-step solvothermal method. HRTEM images showed continuous and well-aligned lattice fringes at the BiOBr/Bi₄O₅Br₂ interface, confirming a high degree of lattice matching, as shown in Figure 8D. These structural features help reduce interfacial charge transfer resistance and provide a foundation for efficient cross-interface migration of photogenerated charge carriers. Theoretical calculations indicated that electrons predominantly migrate from Bi₄O₅Br₂ to BiOBr, resulting in a pronounced charge-density difference at the interface, as shown in Figure 8E. This clearly demonstrates directional electron transfer across the heterojunction. Quantitative charge density analysis along the Z-axis further confirmed electron enrichment on the BiOBr side, as shown in Figure 8F. The BiOBr region displayed increased electron density, while the Bi4O5Br2 area showed a decrease, aligning with the qualitative profile. These findings collectively confirm the direction of electron transfer and the spatial distribution of accumulated electrons. Within the constructed BiOBr/Bi₄O₅Br₂ Z-scheme heterojunction, electrons migrate from Bi₄O₅Br₂ to BiOBr, whereas holes remain in the valence band of Bi₄O₅Br₂. This mechanism promotes the spatial separation of photogenerated electron-hole pairs and effectively inhibits their rapid recombination within individual regions.

In summary, Z-scheme heterojunctions overcome the redox potential loss of Type-II systems by enabling selective recombination of less active carriers, thereby preserving strong reduction and oxidation abilities. Interface engineering - such as conductive bridges, chemical bonding, or noble-metal mediators - is essential to direct the desired charge-transfer pathway and enhance nitrogen-fixation selectivity.

The S-type heterojunction is a semiconductor heterostructure that optimizes the separation and transport of photogenerated charge carriers and is widely used in photocatalysis. Its name comes from the stepped band alignment of two semiconductors, not a literal “S-shape”. When in contact, their energy bands form a staggered configuration for directional charge migration^[102]. Table 3 presents data on S-type heterojunctions over the past few years.

Lv et al.^[61] successfully constructed an S-scheme BiOBr/BiSBr (BOB/BSB) heterojunction featuring a strongly coupled interface with atomic layer bonding (ALB) via an in-situ anion exchange strategy. The interface is characterized by a shared Bi atomic layer and O-Bi-S covalent bonds, which significantly enhance interfacial binding strength and charge-transport capability, as shown in Figure 9A and B. This interfacial engineering strategy drives S-scheme charge transfer and spatial separation via strong orbital hybridization and a reinforced built-in electric field. Meanwhile, oxygen vacancies are introduced to optimize the Fermi level and surface adsorption behavior. The synergistic effect between the strongly coupled ALB interface and oxygen vacancies improves orbital hybridization, promotes N₂ and H₂O activation, and reduces the Gibbs free energy of the rate-determining step, as shown in Figure 9C. Han et al.^[138] precisely constructed a Bi₂O₂CO₃/g-C₃N₄/SrTiO₃ (BOC/CN/STO) dual S-scheme heterojunction via a multi-step hydrothermal process. This substantially lowers the reaction energy barrier by strengthening the built-in electric field, achieving intimate interfacial coupling and strong electronic interactions among the three components, as shown in Figure 9D. Benefiting from dual built-in electric fields and a staggered band alignment, this system drives efficient separation and directional transport of photogenerated charge carriers along dual S-scheme pathways, suppressing interfacial charge recombination. Compared with single components and conventional binary heterojunctions, this interfacial engineering strategy prolongs carrier lifetimes, enhances interfacial charge-transfer kinetics, and preserves the materials’ strong redox capability, as shown in Figure 9E-G.

Recently, our group designed and fabricated a MoS₂/In-Bi₂MoO₆ heterojunction catalyst through electrostatic self-assembly, resulting in interfacial chemical bonding. Instead of mere physical attachment, the Mo-S bonds formed at the interface serve as an “electron bridge” connecting the two materials. This chemical bond, functioning as an electron conduit, removes the interfacial transport barrier. Compared with the Mo-S bond in pure MoS₂^[44], the bond length in the heterojunction is elongated by 0.15 Å, which further facilitates electron transfer, as shown in Figure 10A and B. Meanwhile, the work function mismatch between MoS₂ and In-Bi₂MoO₆ induces an interfacial potential difference, resulting in a built-in electric field that directionally drives the migration of photogenerated carriers, thereby promoting electron transfer from In-Bi₂MoO₆ to MoS₂, as shown in Figure 10C. Thus, the separation of charge carriers is addressed from a thermodynamic perspective, which aligns with the fundamental principles of negative charge carriers to achieve the spontaneous assembly of the two materials, thereby constructing an In₂O₃/Bi₂MoO₆ heterojunction^[58]. At the interface, In-O-Mo chemical bonds form, serving as an “electron bridge” to remove the physical potential barrier, in line with the fundamental principle of interfacial potential regulation in heterojunction systems. Recently, we employed the electrostatic attraction between positively charged Bi₂MoO₆ and the gap to achieve atomic-level intimate contact between the components, as shown in Figure 10D-F. This facilitates electron transfer from In₂O₃ to Bi₂MoO₆ and establishes a continuous pathway for charge transport. The formation of In-O-Mo bonds create a significant number of oxygen vacancies, which enhance nitrogen adsorption and activation, providing more active sites for photocatalytic nitrogen fixation.

More recently, Ren et al.^[133] fabricated an S-scheme Cs₃Bi₂Br₉/BiOBr heterojunction by growing Cs₃Bi₂Br₉ nanocrystals in situ on BiOBr hollow nanotubes. This in situ transformation strategy enables the formation of intimate interfacial contact within the heterojunction. The interface between Cs₃Bi₂Br₉ and BiOBr exhibits a smooth transition without discernible gaps, confirming that the in situ approach achieves atomic-level interfacial intimacy and provides an uninterrupted pathway for charge transfer, as shown in Figure 11A. A strong, homogeneous surface photovoltage signal is observed on the sample surface, indicating efficient separation and surface accumulation of photogenerated charge carriers, as shown in Figure 11B and C. Liu et al.^[134] employed an anti-solvent induction strategy to enable the in-situ growth of Cs₃Mo_xSb_yBr₉ nanocrystals on BiVO₄ nanosheets, thereby forming a tightly connected heterojunction interface. Lattice overlap between the two materials is observed at the interface, without discernible defects or gaps, ensuring efficient charge transport. By varying the amount of MoBr₄ added, a series of Cs₃Mo_xSb_yBr₉/BiVO₄ heterojunctions with different Mo doping levels was synthesized to optimize the interfacial properties. Upon Mo functionalization, the d-band center of Cs₃Mo_xSb_yBr₉ is upshifted toward the Fermi level (from -29.43 eV to -4.31 eV), leading to enhanced N₂ binding and promoted activation of the N≡N bond, as shown in Figure 11D-F.

Recently, our group fabricated a Bi₂S₃/OV-Bi₂MoO₆ heterojunction via an in-situ anion-exchange method^[131], where Bi₂S₃ nanoparticles were uniformly anchored on the surface of Bi₂MoO₆ without observable physical gaps at the interface, achieving intimate interfacial contact. The strong interaction between Bi₂S₃ and Bi₂MoO₆ during interface formation leads to the reduction of Mo⁶⁺ to Mo⁵⁺, accompanied by the generation of abundant OVs, as shown in Figure 11G. These OVs, together with Bi₂S₃, form dual active sites, while the intimate interfacial contact and the introduction of oxygen vacancies synergistically promote N₂ activation, as shown in Figure 11H. Meanwhile, an S-scheme charge transfer pathway enhances the utilization efficiency of photogenerated carriers. Consequently, both the NRR and oxygen evolution reaction (OER) are simultaneously enhanced, as shown in Figure 11I.

Li et al.^[139] constructed a chemically bonded Bi₂Sn₂O₇/Bi₂MoO₆ S-scheme heterojunction. By leveraging the built-in electric field and band bending, they achieved efficient spatial separation of photogenerated charge carriers while preserving strongly reductive electrons and strongly oxidative holes, effectively addressing the key issues of severe carrier recombination and inadequate redox capability in traditional heterojunctions, as shown in Figure 12A. Moreover, the chemically bonded interface induces the formation of asymmetric Bi-Sn dual-atom active sites, enabling side-on adsorption and efficient activation of N₂, thereby significantly reducing the energy barrier for the nitrogen reduction reaction. This enables synergistic regulation of interfacial structure, charge transport, and N₂ activation at the atomic level, ultimately substantially enhancing the photocatalytic nitrogen fixation performance, as shown in Figure 12B.

Zhang et al.^[132] demonstrated the formation of a heterojunction between Bi₂Sn₂O₇ (BSO) and BiOBr (BOB), in which the defective sites (unsaturated Bi and O atoms) on the heterojunction surface interact to create new Bi-O covalent bonds, thereby tightly connecting the two materials, as shown in Figure 12C. This chemically bonded interface, as opposed to mere physical adsorption, establishes a “rigid linkage” that prevents interfacial detachment or charge-transfer blockage. Under illumination, electrons from the CB of BOB and holes from the VB of BSO recombine rapidly via the Bi-O bonds (deactivating less reactive carriers). In contrast, the highly reductive electrons in the CB of BSO and the oxidative holes in the VB of BOB are preserved. The Bi and Sn atoms surrounding oxygen vacancies on the BSO surface provide active sites for N₂ adsorption, as shown in Figure 12D. The accumulated electrons are efficiently transferred through the Bi-O bonds to the adsorbed N₂ molecules, weakening the N≡N bond and lowering the energy barrier to nitrogen reduction.

In summary, S-scheme heterojunctions effectively address key challenges in photocatalytic nitrogen fixation, such as low charge separation efficiency and the difficult activation of N₂, by leveraging stepped band alignment, a built-in electric field, and selective carrier recombination. Further advances in interfacial engineering, active site design, and system optimization may further promote their practical application in green ammonia synthesis.

Beyond the heterojunctions previously discussed, other heterojunctions, such as ohmic heterojunctions, have also been explored. For instance, Zheng et al.^[140] prepared a structurally aligned 2D-2D ohmic heterojunction by epitaxially growing a two-dimensional bismuth layer on BiOBr nanosheets via in situ annealing, as shown in Figure 13A. This process formed a Schottky-barrier-free ohmic contact interface, eliminating the energy barrier to electron transfer and establishing a nearly impedance-free charge-transport channel, as shown in Figure 13B and C. Simultaneously, high-density OVs were introduced at the BiOBr interface during annealing, as shown in Figure 13D. The formed Bi-BiOBr coordination bonds at the interface alter the chemical environment of surrounding atoms, further optimizing the electronic structure for N₂ adsorption, weakening the N≡N bond energy, and promoting the activation and dissociation of N₂, as shown in Figure 13E. In situ Fourier transform infrared Spectroscopy (FTIR) spectroscopy confirmed that the Bi/BiOBr interface significantly enhances the chemical adsorption of N₂, and the adsorbed N₂ can directly participate in the subsequent reactions, as shown in Figure 13F and G.