Metal Sulfide‐Based Nanoarchitectures for Energetic and Environmental Applications

Despite their numerous excellent properties, metal sulfides are not particularly efficient at converting energy and purifying the environment, which limits their further applications. Fortunately, the energy conversion and environmental purification efficiencies of these materials have experienced notable advancements in recent years, accompanied by an improved understanding of their underlying mechanisms. Herein, progress in experimental researches in recent years on the engineering of single component metal sulfides by controlling morphology, construction of heterojunctions, and incorporating elements is reviewed. Methods to design and prepare metal sulfide‐based composites by building binary or ternary heterojunctions of metal sulfide/semiconductor/conductor are also discussed in detail. These materials are used in energy conversion and environmental purification systems, where they act as photocatalytic materials not only to split water, reduce carbon dioxide or nitrogen, but also to degrade pollutants (organic and inorganic) in water and gas. Finally, it is concluded by summarizing the research frontiers of metal sulfide nanomaterials in energy and environmental applications, as well as proposing potential challenges and future research directions. This work may contribute to a better understanding of metal sulfide nanocomposites and provide clues for the fabrication of more efficient metal sulfide‐based nanostructures for clean energy production and environmental remediation.


Introduction
With the ongoing advancement of industrial technology, we are witnessing the prosperity and progress of this era.However, the earth is at risk due to environmental pollution and energy shortages. [1]The discharge of industrial wastewater causes to the continuous accumulation of pollutants in water bodies.Accordingly, aquatic ecosystems that are crucial to human and animal life have suffered varying degrees of damage, making available water resources continuously reduced.Pollutants in wastewater can be divided into inorganic pollutants and organic matter based on their composition.Common inorganic pollutants include numerous heavy metals (such as copper, arsenic, and lead), [2] ammonia nitrogen, phosphorus, sulfides, and fluorides.[5] People can suffer serious health hazards and unimaginable consequences when drinking or indirectly using water that contains these substances.Traditional organic wastewater treatment technologies are grouped mainly into physical, chemical, and biological methods. [6,7]10][11] The commonly used chemical methods, such as Fenton catalytic reactions and Fentonlike catalytic reactions, require acidic conditions or a large amount of H 2 O 2 to work well. [12]Biological methods have the potential to eliminate pollutants, but some organic compounds cannot be effectively mineralized due to their complex structure, which produces more toxic compounds and secondary pollution. [10]Therefore, it is necessary to develop a new type of organic wastewater treatment technology that is green, efficient, low energy consumption, and capable of deep oxidation and decomposition of refractory pollutants.
As an advanced oxidation process, photocatalytic technologies gained rapid momentum due to Fujishima and Honda's discovery of photoelectrochemical water splitting. [13]In recent years, various semiconductor photocatalysts have been widely used in environmental remediation and clean energy conversion, such as the reduction of CO 2 , the decomposition of water, and the treatment of wastewater, which are both economical and environmentally friendly. [14,15]These use of photocatalysts driven by visible light has been developed to make solar energy more efficient.
In recent years, oxide semiconductors such as TiO 2 , [16,17] ZnO, [18,19] SnO 2 , [20] and WO 3 [21] have been studied, which are used to solve environmental problems and have achieved some success.[24] However, TiO 2 cannot be excited by light other than UV light because of its wide bandgap, which severely limits its application. [23,25]ZnO has low cost, good stability, and high catalytic activity, but photocorrosion will occur during the photocatalytic process, resulting in reduced catalytic activity.SnO 2 can only excite and generate electron-hole pairs under UV light irradiation, and the photogenerated electron-hole pairs will quickly recombine.[28] The inappropriate intrinsic band structure of these metal oxide photocatalysts limits their further application.
During the last few years, metal sulfides containing sulfur element in the same main group (VIA) as oxygen element have gained increasing attention from researchers, such as binary metal sulfides (CuS, ZnS, CdS, FeS 2 , MoS 2 ), ternary metal sulfides (ZnIn 2 S 4 , CuInS 8 , CdIn 2 S 4 , CuFeS 2 ), and quaternary metal sulfide (Cu 2 ZnSnS 4, Cu 2 FeSnS 4 , Cu 2 CoSnS 4 ).An overview of typical metal sulfide photocatalysts in the last 10 years is presented in Table 1.Metal sulfides semiconductor photocatalysts have more suitable band structure, allowing them to exhibit better light absorption and effective oxidation or reduction capabilities over a wide wavelength range. [29]But like metal oxides, they also have problems such as unstable band structure and large thermal effects, leading to a large number of improvement measures being implemented to optimize their photocatalytic performance.Here, we examine the latest advancements made in the development and utilization of visible light responsive metal sulfide nanostructures, including binary, ternary, and quaternary sulfides.The high-temperature solid-phase method, hydrothermal (solvothermal) approach, chemical vapor deposition (CVD), sol-gel method, electrospinning, biosynthesis method, and other solid-, liquid-, and gas-phase preparation methods are first summarized.The strategies to optimize the catalytic performance of metal sulfide nanostructures are subsequently discussed, mainly focusing on controlling the morphology, doping, and constructing heterojunction structure.The internal structure and reaction mechanism of these metal sulfide-based photocatalysts are also elaborated in detail.Furthermore, different metal sulfide semiconductors are systematically scrutinized to develop environmentally friendly and efficient methods for pollutant treatment and energy conversion applications under visible light irradiation, including photocatalytic pollutant degradation, CO 2 reduction, water splitting, and nitrogen reduction.

Synthesis Methods for Metal Sulfur-Based Semiconductors
The performance and ultimate cost of photocatalysts are two key factors that need to be taken into account in industrial applications, which are intricately linked to each meticulously chosen stage, including the synthesis process, raw material selection, reaction condition control, and equipment utilization.Therefore, appropriate synthesis methods are particularly important in achieving efficient and environmentally friendly photocatalyst production.We have summarized some common preparation methods for metal sulfide-based photocatalysts (Figure 1), including solid-phase method, liquid-phase method, gas-phase method, electroplating method, microwave-assisted method, biosynthesis method, and so on.The merits and drawbacks of these methods have been assessed in terms of raw materials, methodologies, equipment, and reaction conditions.

Solid-Phase Synthesis
Solid-phase method is a process that reduces the size of solidphase materials or recombines them by means of grinding or heat treatment of raw materials, [30] including high-energy ball milling, solid-phase reaction, thermal decomposition, and spark discharge approach.Park et al. prepared ZnS/C nanocomposites by ball milling (Figure 2a). [31]The composites were composed of ZnS nanocrystals with a particle size of 5-10 nm distributed on the amorphous carbon matrix (Figure 2b).Li et al. synthesized Cu 2 MoS 4 nanoparticles at 200 °C for 3 h by a solid-state reaction (Figure 2c). [32]The microstructure of Cu 2 MoS 4 synthesized without the addition of surfactant is composed of a large number of small and thick nanoparticles with obvious and irregular aggregation.Aiganym et al. pretreated the Zn source precursor with NaCl as diluent and then milled the precursor in a planetary ball mill, after which the prepared NaCl-based ZnS nanoparticles were calcined at 500 °C with controlled heating duration (1, 2, 3, and 4 h) to obtain ZnS/ZnO nanoparticles. [33]It was found that the size of the nanoparticles was observed to change with the change in heating time.Scanning electron microscopy studies showed that the size of individual nanoparticles of ZnS/ZnO with a heating duration of 1 h ranged from 10 to 150 nm with significant aggregation and no uniform distribution.The particle size was slightly thickened with increasing time (Figure 2d).Compared with the ball milling method, the product synthesized by the high-temperature solid-phase method usually has higher purity and more uniform grain size.In addition, the use of different reaction temperatures or reaction times in the solid-phase reaction can lead to changes in the size of the nanoparticles, which can affect the photocatalytic performance.Solid-phase method is time-saving, environmentally friendly, high yield, easy to prepare, and appropriate for industrial production.However, because the reaction materials are solid, it is difficult to diffuse between the reactants, which will lead to the uneven morphology of the synthesized nanomaterials.For this reason, in recent years, researchers have rarely used the solid-phase method to synthesize nanomaterials alone, and often combine the solidphase method with other methods to synthesize nanoparticles with relatively uniform morphology and size.

Hydrothermal (Solvothermal) Approach
Hydrothermal (solvothermal) synthesis approach involves maintaining a certain temperature and pressure as well as dissolving the metal salts and sulfur source completely in the reaction solvent.By controlling the temperature and time of the reaction, different forms of metal sulfide crystals can be obtained.Gou and co-workers synthesized Znln 2 S 4 nanotubes (Figure 3b,c) and nanobelts (Figure 3d,e) with pyridine as solvent (Figure 3a).Interestingly, adding an appropriate amount of surfactant to the hydrothermal (solvothermal) synthesis reaction helps to form a variety of microscopic features.As shown in Figure 3f,g, Znln 2 S 4 solid and hollow microspheres were obtained when surfactants such as cetyltrimethylammonium bromide or polyethylene glycol were added. [34,35] 3h,i) by hydrothermal method by simply changing the metal precursor and adding surfactant in the process of structure formation. [36]It was found that the metal precursor affects the final Znln 2 S 4 polymorph, that is, cubic Znln 2 S 4 is formed when metal nitrate is added, while hexagonal phase is formed when metal chloride is added.Yu et al. used an aging reaction mixture including amine, sulfur, and diethylzinc to synthesize quasispherical and rod-like quantum-sized zinc sulfide nanocrystals. [37]Based on the aforementioned analysis, it is evident that the hydrothermal (solvothermal) method can easily adjust the crystal structure, microstructure, and grain size of photocatalysts through the manipulation of the reaction temperature, time, precursor concentration, or surfactant type.Consequently, this method exhibits promising potential for application in the domains of environment and energy.In comparison, hydrothermal (solvothermal) methods often necessitate a lower temperature (≈200 °C) or more cost-effective raw materials to product fully developed, high-purity, and uniformly sized photocatalysts at the micrometer to nanometer scale.In spite of this, this method requires the synthesis of materials under high pressure and time-consuming preparation process, which involves expensive equipment and complicated technology.

Sol-Gel Method
At low temperatures, sol-gel approach produces organicinorganic composites by hydrolyzing constituent molecular precursors and subsequently polycondensing them into jelly-like structures.Preparation techniques mainly involve templatebased, surfactant-directed or interfacial approaches.In contrast, Zhou and co-workers prepared uniform and undersized 10 nm hexagonal copper sulfide nanorods by the gel-solution method without using templates and complex postprocessing, which reduces the difficulty and cost of the operation and avoids the contamination generated by the sintering process. [38]The sol-gel method can be used to improve the catalytic activity of  Reproduced with permission. [31]Copyright 2018, Elsevier.c) Schematic illustration of high temperature solid-phase reaction.d) scanning electron microscope (SEM) images of ZnS/ZnO at different reaction times.Reproduced with permission. [33]Copyright 2023, Elsevier.
photocatalysts by adjusting parameters such as precursor concentration, temperature, reaction time, and so on to control the size, distribution, and special structure of the particles.Sankar et al. successfully synthesized Ni-doped CdS nanoparticles with different doping percentages by sol-gel method. [39]oping achieved by sol-gel method to obtain photocatalysts with small grain size (average crystallite size of 8.70-9.93nm) and uniform size seems to be a significant advantage.In addition, sol-gel methods can also be well integrated with other preparation methods for the desired multiple heterojunction.Nasseh et al. synthesized magnetic nanocomposites CuS/Fe 3 O 4 /GO consisting of copper sulfide, iron oxide, and graphene oxide by replacing 12% Cu with equal amounts of Fe 3 O 4 and GO through a two-step sol-gel and hydrothermal method. [40]The sol-gel method burns off the organic matter in the sample directly by heating to obtain the final product.This method reduces the mass loss in the recovery process and minimizes the waste of resources compared to traditional methods such as filtration, centrifugation and magnetic separation of nanocomposites.In summary, the sol-gel approach offers the advantage of operating at a reduced temperature (≈100 °C) and achieving more precise control over the photocatalyst structure due to the uniform dispersion of raw ingredients.However, the process of preparing sol-gel takes a long time, ranging from a few days to a few weeks.More importantly, due to the involvement of ester compounds, metal alcohol salts, and other organic raw materials, the raw materials used are expensive, and certain toxic and harmful gases or organic substances may escape during the gel drying process, which will not only cause shrinkage of the gel volume, but also jeopardize human health.

CVD
CVD is a method by evaporating the required compounds, causing them to undergo chemical reactions in the gas phase, followed by fast condensation within a protective gas environment.In recent years, CVD is often used in the preparation of films, powders, fibers, blocks, and other materials.Samad and co-workers fabricated phase-pure pyrite films by CVD using FeCl 3 and di-tert-butyl disulfide as precursors and cordierite CoS 2 as substrate. [41]Moon et al. obtained premium ZnS nanowires with a wurtzite structure with a diameter of 10-30 nm and a length of micrometer scale along the [001] direction by CVD. [42]u and co-workers used CVD method to grow CuS in situ on Cu 2 O to form a core-shell structure, which can effectively inhibit the photocorrosion of Cu 2 O. [43] Additionally, the core-shell structure of Cu 2 O@CuS can be expanded to increase the light absorption range of Cu 2 O from 200 to 1500 nm to enhance visible light catalysis and stability.CVD can produce nanomaterials with high purity, uniform distribution, and good formability, but it requires complex and expensive reactors or vacuum systems, and the process is complicated and the operating conditions are harsh.Reproduced with permission. [34]Copyright 2006, American Chemical Society.h,i) SEM images of Znln 2 S 4 cubic NPs and hexagonal FMs.Reproduced with permission. [36]Copyright 2011, Royal Society of Chemistry.

Electrospinning Method
During electrospinning, polymer fluids are split electrostatically, allowing for the formation of tiny jets of polymers, which can run for an extended distance and then solidify into fibers.In Figure 4a, a schematic illustration depicts the electrospinning device, which consists primarily of three parts: and a collector (usually a rotating mandrel or a metal filter), a syringe with a metal needle, and a high-voltage power supply. [44]According to the research of Zhou et al. zinc ions and sulfur ions were successively added to carboxylated polyvinyl alcohol (PVA) nanofibers to form ZnS nanoparticles by an electrospinning device. [45] typical PVA nanofiber is 300 nm in diameter, and the typical ZnS nanoparticle is 5 nm in diameter (Figure 4b).By electrospinning, not only solid nanofibers but also hollow nanotubes can be obtained.Nakamoto's team combined the elemental sulfur melt electrospinning method with the electroless copper plating technology.[46] Using the electrospun sulfur fiber as a template, the electrospun fiber was immersed in an electroless copper plating bath to prepare various micron-sized copper sulfide (CuS) tubes (Figure 4c,d).Sung and co-workers also embedded zinc sulfide nanoparticles inside a unique tunnel mesoporous carbon fiber and obtained a hybrid ZnS/SCNF-T3 nanocomposite (Figure 4e,f ).[47] Electrospinning has become a popular method for preparing 1D nanofibers and tubular nanomaterials due to its inherent benefits, including its ease of operation, controllability, and large range of spinnable materials.Nevertheless, this process necessitates the utilization of intricate apparatus and additional high-temperature heating systems, and is subject to electrostatic interference.

Microwave-Assisted Method
A microwave is capable of penetrating materials and transferring energy for heating relatively thick materials quickly and uniformly.When a microwave is in contact with materials with different dielectric properties, new or unique microstructure materials with particular properties can be developed by selectively heating the reactants.Chachvalvutikul et al. prepared Bi 2 WO 6 /ZnIn 2 S 4 nanocomposites with different mass fractions by cyclic microwave irradiation combined with wet impregnation method. [48]The photocatalyst still maintains high photocatalytic activity after three repeated cycles, which can degrade methylene blue and produce H 2 .Using the microwave method, Wang and co-workers prepared pure ZnS and different metal ion-doped ZnS catalysts.Cu:ZnS demonstrates the highest degradation rate for 4-chlorophenol (4-CP) during the heterogeneous photocatalytic ozonation process. [49]Wang et al. synthesized a series of g-C 3 N 4 -loaded and Fe-doped ZnS (Zn 1-1.5xFe x S/g-C 3 N 4 ) composite catalysts by a rapid microwave-assisted method. [43]By depositing Fe:ZnS on g-C 3 N 4 surface, Fe:ZnS aggregation can be effectively prevented and surface area can be increased, which promotes mass transport and light absorption.Presently, microwaves are used to assist other methods to synthesize materials.Overall, the microwave heating method is uniform and fast, Reproduced with permission. [45]environmentally safe, and energy-saving.However, it is crucial to use a pure material to prevent any reactions with impurities and the resulting material contamination.

Biosynthesis Method
Biosynthesis refers to the use of viable organisms, such as bacteria, fungi, viruses, animals, and plants, to synthesize inorganic nanoparticles at room temperature and near neutral pH under mild conditions.It is a clean, nontoxic, and pollution-free green synthesis method. [50]Qin and co-workers used biomineralization as a template and stabilizer to synthesize silver-doped copper sulfide (Ag:CuS) nanoparticles with strong antitumor ability, good biocompatibility, low toxicity, and good stability by using bovine serum albumin. [51]Ag:CuS absorb near-infrared light more efficiently and are more photothermally efficient.In addition to improving the therapeutic efficacy of photothermal chemical kinetic therapy, this method also provides therapeutic synergy to overcome poor biocompatibility, low photothermal efficiency, and harsh preparation conditions.Samanta et al. used the biomolecule (lemon juice)-assisted preparation scheme for tailoring the morphology of synthesized ZnS nanostructures (nanospheres, nanoflowers, and nanotubes) by changing the ratio of precursors and stabilizers. [52]ZnS nanospheres have high separation efficiency and surface-to-volume ratio of electron-hole pairing, and the photodegradation rate of rhodamine B (RhB) utilizing sunlight is 85%, and the degradation rate of methyl orange (MO) is 82%.Yue et al. achieved the controllable synthesis of wurtzite (α-ZnS) and sphalerite (β-ZnS) quantum dots by synthesizing ZnS using mixture of clostridium, hydroxypropyl starch, EDTA-Zn, and Na 2 SO 4 . [53]The diameter of α-ZnS quantum dots synthesized by low-dose hydroxypropyl starch was 6-10 nm, while the diameter of β-ZnS quantum dots synthesized by high-dose hydroxypropyl starch was 3-5 nm.Biosynthesis has the advantages of high yield, low cost, good biocompatibility, high stability, and narrow size distribution.In spite of this, the biosynthesis method is not yet clear about the mechanism of synthesis, and the synthesis process is relatively difficult to control and time-consuming.

Design Strategy of Metal Sulfur-Based Semiconductors
Efficient light absorption and strong redox ability are the basic prerequisites for photocatalysis. [54]However, practical applications of photocatalysis are limited by individual nanomaterials, which often have weak selectivity for reaction products and fast recombination of photogenerated charge carriers. [55]In addition to choosing the suitable photocatalysts and synthetic methods, structural control strategies also play a crucial role in designing efficient photocatalytic materials. [56]Common structural control methods include element doping, [57] morphology control, [58] defect engineering, [57] and heterostructure building. [59]

Morphology Control
Different morphologies and structures affect the photocatalytic performance of semiconductor photocatalysts.As photocatalysis takes place predominantly on semiconductor material's surface, the regulation of the morphology can increase the specific surface area of the material, enhance the exposure of the active site, [58,60] promote light absorption and energy transmission, [61] accelerate charge transfer, [62] and manipulate the separation and recombination of photogenerated electron-hole pairs. [63]ccordingly, researchers have synthesized semiconductor photocatalysts with all kinds of morphologies to boost photocatalytic performance.Figure 5a-i illustrates semiconductor photocatalysts with different morphologies.There are 0D nanostructures, composed of very small particles, without large anisotropic aspect ratio, [64] including quantum dots, nanoparticles, and so on.1D usually has nanorods, nanowires, nanofibers, nanowhiskers, and nanotubes.The photogenerated electrons on nanostructures such as nanotubes can move rapidly along the 1D axis, reducing the recombination of carriers. [65]Photocatalysts with 2D structures include nanosheets, nanobelts, nanosheets, films, and so on.This kind of structure reduces the thickness of the material, shortens the distance from the photogenerated electrons to the surface, and enhances the utilization rate of electrons.Moreover, the lamellar structure can expose more reactive sites, which greatly affects the photocatalytic efficiency. [66,67]3D mesoporous structure, core-shell nanoparticles, hollow nanospheres, nanoflowers, layered nanostructures, and so on increase the light absorption area, produce more light reflection and scattering, and have excellent light absorption capacity.Meanwhile, with the 3D spatial structure, photogenerated carriers recombinate less frequently and photocatalysis performance is further enhanced. [68]

Element Doping
In semiconductors, element doping plays an important role in improving photocatalytic activity and can help improve charge-hole separation.Two common types of element doping are metal doping and nonmetal doping. [69,70]With element doping, the semiconductor type and carrier concentration can be changed, and vacancies can act as electron traps, preventing carriers from recombinating, thereby prolonging the lifetime of photogenerated electrons. [71]The element doping can generate new acceptor energy levels below the conduction band (CB) or above the valence band (VB) of the semiconductor, reduce the original bandgap width of the semiconductor, expand the absorption region of visible light, and finally improve the photocatalytic activity. [72]

Metal Element Doping
Metal element doping involves introducing metal elements into the material body through a certain chemical preparation method to replace the crystal lattice atoms of the host material.The doping of impurity elements can be used to adjust photogenerated charge separation rate and electronic structure of the photocatalyst, thereby improving the ability to absorb visible light and the synergistic effects of effective charge carrier separations. [73]Jia and co-workers obtained La-doped ZnIn 2 S 4 microspheres by a one-pot hydrothermal synthesis process, and the samples showed a significant enhancement in the photodegradation of tetracycline hydrochloride and MO. [74]Venkatesh et al. used polyvinylpyrrolidone as capping agent to prepare Sn-doped CdS nanoparticles (Sn:CdS) with various doping amounts (1-5% and 10%) by chemical precipitation method. [75]Optical bandgap for CdS was adjusted by different Sn concentrations, and the photocatalytic degradation rate of 4% Sn:CdS nanoparticles was 97.56% after 180 min.A semiconductor material will generate a new VB position higher than the original VB position or a new CB position lower than the original CB position when metal ions are doped into it (Figure 6, left).The bandgap of Cu: ZnS nanoparticles was tuned from 3.93 to 2.72 eV, and the surface area was increased from 89.5 to 152.7 m 2 .Cu 2þ serves as an electron capture center, achieving a reduction in electron and hole recombination.Under natural illumination, photocatalytic experiments on 4-CP degradation showed 69.2% higher efficiency than that of undoped ZnS nanoparticles. [76]As reported by Gao et al.Fe-doped ZnIn 2 S 4 photocatalysts are effective in improving the photoactivity of halogenated compounds and in removing 2,4,6-tribromophenols. [77]an and co-workers prepared some rare earth (RE) ion-doped ZnIn 2 S 4 photocatalysts (RE:ZnIn 2 S 4 , RE = La 3þ , Ce 3þ , Gd 3þ , Er 3þ , or Y 3þ ) utilizing hydrothermal method and characterized their properties. [78]They show that the RE elements exist in the form of oxide RE 2 O 3 , which can reduce the grain size of ZnIn 2 S 4 , inhibit the grain growth of ZnIn 2 S 4 , promote the crystal self-organization of ZnIn 2 S 4 to form a microsphere flower-like structure, boost the total pore volume and surface area of ZnIn 2 S 4 , and make ZnIn 2 S 4 produce rich defects.After adding 2.0% of Y, Gd, Er, Ce, and La, the hydrogen production efficiency increased by 46%, 53%, 61%, 69%, and 106%, respectively.

Nonmetallic Ion Doping
In contrast to metal ion doping, nonmetal element doping can form a donor energy level in the forbidden band.The VB edge can be moved upward due to the impurity state of the anion, l) Hierarchical nanostructures.Reproduced with permission. [64]Copyright 2016, American Chemical Society.
causing a change in the bandgap of the semiconductor catalyst (Figure 6, right), so as to achieve the modification effect of the photocatalyst.Ma et al. used local density approximation theory to study the p-type doping properties of ZnS nanocrystals. [79]It was found that the systems doped with only one element of N, P, or As exhibited lower photocatalytic activity than the p-type ZnS nanocrystal composites with high doping density and high doping efficiency (codoped with N, Ga, In, and Al).This is mainly attributed to the lower doping concentration and efficiency of the ZnS nanocrystals, which is limited by the expulsion effect between Zn and impurity atoms as well as the compensation effect provided by interstitial atoms.Liu et al. synthesized N-doped MoS 2 nanoflowers via sol-gel approach, which have wonderful photocatalytic activity in the degradation of RhB and other organic dyes and heavy metals. [80]This investigation found that there are few studies on nonmetallic doping of metal sulfide semiconductors, which may be due to the relatively narrow bandgap of metal sulfides.After doping of nonmetallic elements, the light response range and CB position will be affected, thus reducing the oxidation potential and preventing the redox reaction from proceeding.However, there are relatively many studies on the doping of metal sulfide heterojunctions by nonmetallic elements.Some scholars have investigated the degradation of MO by N-doped and CdS-coupled TiO 2 photocatalysts under visible light.CdS/N-TiO 2 photodegrades of MO extremely well, exhibiting many times the activity of N-TiO 2 and CdS/TiO 2 in the presence of visible light. [81]Hu and co-workers designed and prepared a particular hierarchical CNFs@I-doped Bi 2 O 2 CO 3 -MoS 2 carbon nanofiber film.The introduction of I replaces some CO 3 2À , which narrows the bandgap of Bi 2 O 2 CO 3 and modifies the surface of CNFs. [82]The introduction of MoS 2 nanosheets is more conducive to electron transport and prevents the rapid recombination of photogenerated electron-hole pairs.The nanocomposites of CNFs@I-doped Bi 2 O 2 CO 3 -MoS 2 can effectively photodegrade wastewater pollutants including RhB.

Heterojunction Construction
It is effective to improve the separation between photoelectron and hole by means of heterojunction engineering.A variety of materials can be used to combine metal sulfide semiconductors to form heterojunctions, for instance, type II (Figure 7a), p-n (Figure 7b), Z-scheme (Figure 7c), and S-scheme (Figure 7d) heterojunctions.

Type II Heterojunction
Among the heterojunctions used in photocatalysis, type II heterojunction with 1D/1D, 0D/1D, and 1D/2D structures is the most common.In the type II heterojunction, as a result of light irradiation, the electrons generated in CB of semiconductor 1 are transferred to CB of semiconductor 2, while semiconductor 2 transmits photogenetic holes to semiconductor 1 through its VB. [83]An internal electric field (IEF) is formed at the interface between semiconductor 1 and semiconductor 2, so that the photogenerated electron-hole pairs can be effectively separated, and the photocatalytic activity of the semiconductor photocatalyst can be significantly improved.Zhang et al. synthesized a 1D/1D core/shell nanowires coated with g-C 3 N 4 on the surface of CdS nanowires. [84]CdS is capable of easily transferring holes from its VB to its high occupied molecular orbital by excitation with visible light, and the photogenic electrons on the molecular orbital of g-C 3 N 4 are easily injected into its CB (Figure 8a-c).CdS/g-C 3 N 4 has an integrated band structure and close interface contact because of the synergistic effect between g-C 3 N 4 and CdS, which greatly improves the photosensitivity and effectively accelerates the separation/transfer of charge carriers, making the photocatalytic activity enhanced.
Cui et al. successfully constructed a Bi 2 MoO 6 /In 2 S 3 type II heterostructure by growing In 2 S 3 particles on the surface of Bi 2 MoO 6 microspheres using a hydrothermal approach. [85]ffected by the heterostructure, the e À on the lowest unoccupied Hua and co-workers prepared a 2D La 2 Ti 2 O 7 /In 2 S 3 type II heterojunction by a simple self-assembly method. [86]The photogenerated electrons in In 2 S 3 are driven to La 2 Ti 2 O 7 by an electric field and then participate in the water reduction reaction (Figure 8e).The unique face-to-face contact of In 2 S 3 and La 2 Ti 2 O 7 nanosheets in the heterojunction makes it easy for electrons to migrate from In 2 S 3 to La 2 Ti 2 O 7 (Figure 8f,g), thereby promoting efficient charge separation and excellent photocatalytic activity.To improve photocatalytic and photoelectrocatalysis (PEC) performance, other similar type II heterojunctions such as CdS/ZnO, CdS/MoS 2 , ZnIn 2 S 4 /BiPO 4 , and TiO 2 /MoS 2 can be used.

p-n Heterojunction
In a p-n type heterojunction, electrons diffuse through the p-n junction interface from the n-type semiconductor to the p-type semiconductor before light irradiation, owing to the combination of energy band orientation and IEF.Simultaneously, holes in p-type semiconductors incline to spread into n-type semiconductors. [87]As electrons and holes diffuse through the junction, an IEF forms at the interface of the p-n junction, thus achieving Fermi-level equilibrium. [88]With the synergistic effect of IEF and light irradiation, holes and electrons transfer directionally, inhibiting electron recombination and further improving photocatalytic efficiency.
Mondal and co-workers prepared ZnS nanoflowers by modified hydrothermal technology, and mixed them with CuSO 4 aqueous solution to obtain CuS/ZnS photocatalyst. [89]In comparison with pure ZnS and CuS, visible-light photocatalytic performance is significantly strengthened in the CuS/ZnS heterojunction due to the effective electron-hole separation at their interface, which is due to the difference between the p-type CuS and n-type ZnS semiconductors (Figure 9a).Gholami et al. synthesized a visible light driven CuFeS 2 /CuS (CFC) p-n heterojunction by using CuS (p-type) as a carrier and matching with n-type CuFeS 2 semiconductor. [90]The semiconductor with matching energy band structure can provide extra IEF (Figure 9b); electrons and holes are transferred more quickly through a heterojunction, reducing the chances of electron-hole pairs recombination, and accelerating electron separation from photoinduced electrons, and increasing the yield of active substances (•O 2 À , h þ and •OH), which shows high photocatalytic activity.Malathion (25 mg L À1 ) in aqueous solution can be completely degraded within 80 min.
Consequently, compared with the type II heterojunction, the p-n heterojunction is more effective in separating electron-hole pairs and solving the problem of rapid recombination of photogenerated carriers.Further, it enhances the absorption of solar wavelengths and extends the range of visible light absorption edges.However, there is inevitably the problem of weak redox ability.

Z-Scheme Heterojunctions
A Z-scheme heterojunction participates in redox reactions by combining photogenic electrons from semiconductor 1 with photogenic holes from semiconductor 2, which are placed relatively high in semiconductor 1 and relatively low in semiconductor 2, which facilitates redox reactions.Fu and co-workers realized the transformation of ZnO/MoS 2 nanoarrays from type II heterojunction to Z-scheme heterojunction on nickel foam by flow-induced piezoelectric field. [91]Under sunlight irradiation, nanoarrays of ZnO/MoS 2 grew on nickel foam degrade MO.The degradation rate of MO by stirring induced water flow reached 92.7%, which was 42.1% higher than that without stirring.This is mainly attributed to the fact that the barrier height between the CB layer of ZnO and the CB layer of MoS 2 is significantly reduced under the action of the piezoelectric field, which makes it easier for the electrons in the CB layer of ZnO to recombine with the holes in the VB layer of MoS 2 , thus forming a representative Z-scheme heterojunction.As shown in Figure 10a,b, photogenerated electron-hole pairs migrate and reactive oxygen species (ROS) are generated under stirring and without stirring, respectively.Under the irradiation of sunlight, the holes in the VB of MoS 2 combine with the photoelectrons in the CB of ZnO, leaving the holes in the CB of ZnO and the photoelectrons in the VB of MoS 2 .As a result, the photocatalyst's redox ability and activity are enhanced.
Ouyang and co-workers synthesized ZnIn 2 S 4 @MoO 3 direct Z-scheme heterojunction with layered core-shell structure toward visible-light degradation of the main pollutant tetracycline hydrochloride in water. [92]The high photocatalytic activity of ZnIn 2 S 4 @MoO 3 photocatalyst under visible light is due to the Z-scheme photogenerated carrier transfer mechanism (Figure 10c).Because of the built-in electric field force, electrons in the MoO 3 's CB and holes in the ZnIn 2 S 4 's VB combine straight when irradiated by visible light.The VB holes in  [84] Copyright 2013, American Chemical Society.d) Photocatalytic mechanism for degrading RhB by In 2 S 3 /Bi 2 MoO 6 .Reproduced with permission. [85]Copyright 2020, Elsevier.e) Schematic illustration of the band edge positions for both La 2 Ti 2 O 7 and In 2 S 3 .f,g) Proposed band structure and photocatalytic reactions for La 2 Ti 2 O 7 /In 2 S 3 nanosheet heterojunctions.Reproduced with permission. [86]Copyright 2019, Elsevier.
MoO 3 and the CB electrons in ZnIn 2 S 4 remain intact, resulting in electrons and holes being separated from each other, and stronger redox ability and prolonged lifetime.
In addition to the spontaneous formation of Z-scheme heterojunctions between semiconductors, the formation of Z-scheme heterojunctions can also be promoted by adding sacrificial agents.Hao et al. prepared CdS/ZnS photocatalyst with Znvacancy in situ by using CdS/MOF-5 as a sacrificial agent along with Na 2 SO 3 and Na 2 S solution. [93]As shown in Figure 10d, in the presence of light, the photoinduced electrons of ZnS recombine with the holes in CdS through ohmic contact, and then get transferred to the zinc vacancy defect level of ZnS.Inhibiting photocorrosion can be achieved by effectively separating holes and electrons in CdS's CB and VB.Therefore, proton H þ can be reduced to H 2 using photogenerated electrons in CdS's CB.Also, two-photon absorption can be used to transfer electrons from Zn-deficient ZnS to CdS in the CB.At the same time, the VB holes of ZnS are expended by the electron donors Na 2 S and Na 2 SO 3 , which makes the CdS/ZnS heterojunction have high stability and excellent photocatalytic activity.
Z-scheme heterojunctions that can provide efficient photocatalytic activity are widely favored by researchers.However, due to the variety of types and structures of heterojunctions, it is impossible to visually see the overall configuration.The data obtained by the test may also not be able to accurately determine what type of semiconductors are formed.In general, active sites and photocatalytic properties can be combined and their configurations can be determined by hypothesis backstepping.Adil et al. formed a Cu 2 ZnSnS 4 /Pt/g-C 3 N 4 heterojunction photocatalyst with a cross-band configuration for the reduction of carbon dioxide to carbonaceous fuel. [94]In order to obtain the heterojunction type of Cu 2 ZnSnS 4 /Pt/g-C 3 N 4 , the author proposes that the transfer mechanism of interface electrons may be "type I heterojunction transfer" or "Z-scheme transfer".The energy band structure of Cu 2 ZnSnS 4 /Pt/g-C 3 N 4 heterojunction nanostructure and mechanism of CO 2 photoreduction are shown in Figure 10e,f.In the "type I heterojunction transfer," the photoinduced electron-hole pairs in CB and VB of g-C 3 N 4 are transferred to CB and VB of Cu 2 ZnSnS 4 under visible light irradiation, and CO 2 reduction and H 2 O oxidation occur in CB and VB of Cu 2 ZnSnS 4 , respectively.In addition, the CB of Cu 2 ZnSnS 4 is lower than the negative redox potential of CO, and only CH 4 is not generated.However, under visible light irradiation, all photoinduced electron-hole pairs will aggregate on Cu 2 ZnSnS 4 in the form of "type I heterojunction," which not only fails to improve charge separation, but also limits photocatalytic efficiency, which is contrary to the fact.Therefore, it is concluded that the heterojunction type of Cu 2 ZnSnS 4 /Pt/ g-C 3 N 4 belongs to the Z-scheme transfer mechanism, which can better reflect the high catalytic activity and is more suitable for explaining the improvement of solar fuel conversion rate.
It can be seen from the above examples that the Z-scheme heterojunction can effectively improve the carrier separation efficiency, expand the light response range, and retain higher redox active sites, which has great potential in photocatalytic applications.

S-Scheme Heterojunction
Unlike Z-scheme heterojunctions, S-scheme heterojunctions have higher redox abilities and are better at separating and transferring photoexcited carriers.It is composed of two n-type Reproduced with permission. [91]Copyright 2020, Elsevier.c) Charge transfer mechanism of Z-scheme ZnIn 2 S 4 @MoO 3 heterojunction.Reproduced with permission. [92]Copyright 2021, Elsevier.d) Type II CdS/MOF-5, Z-scheme Zn-defect CdS/ZnS heterojunction under visible light irradiation.Reproduced with permission. [93]Copyright 2018, Elsevier.e) Band structure of (left) g-C 3 N 4 and Cu 2 ZnSnS 4 (right) 3 wt% Cu 2 ZnSnS 4 /Pt/g-C 3 N 4 heterojunction photocatalyst.f ) Schematic illustration of CO 2 reduction mechanism for Cu 2 ZnSnS 4 /Pt/g-C 3 N 4 .Reproduced with permission. [94]Copyright 2020, Elsevier.semiconductors 1 and 2 in the S-scheme heterojunction.The former is an oxidation-type photocatalyst with a large work function and low Fermi level, while the latter is a reduction-type photocatalyst, whose work function and Fermi level are opposite.The photoexcited charge transfer pathway in the S-scheme heterojunction is more like a "step" or "N," which means that the holes from the VB of semiconductor 1 and semiconductor 2 are recombined; afterward, oxidative and reduction-capable electrons and holes are spatially detached.
Zhou and co-workers grew ZnIn 2 S 4 nanosheets on the surface of flaky Bi 4 Ti 3 O 12 by low-temperature solvothermal method, and synthesized ZnIn 2 S 4 /Bi 4 Ti 3 O 12 heterojunction with 2D/2D structural characteristics. [95]The researchers found that the energy structures of ZnIn 2 S 4 and Bi 4 Ti 3 O 12 are staggered (Figure 11a).By recombining the electrons in the CB of ZnIn 2 S 4 and the holes in the VB of Bi 4 Ti 3 O 12 , the S-scheme ZnIn 2 S 4 /Bi 4 Ti 3 O 12 heterojunction retains the useful holes and electrons of the two semiconductors.Hence, the charge transport mechanism of S-scheme heterojunction makes ZnIn 2 S 4 and Bi 4 Ti 3 O 12 have higher redox ability to degrade TC.A rate constant of almost 6.8 times the primitive ZnIn 2 S 4 was achieved for the optimized heterojunction when exposed to visible light that allows removal of 82.1% of TC within 60 min.
Su and co-workers calculated the S-scheme heterojunction in CdS quantum dots/Bi 2 WO 6 monolayer system by density functional theory, and then combined CdS quantum dots with Bi 2 WO 6 monolayer system by in situ hydrothermal synthesis method. [96]The band structures of CdS and Bi 2 WO 6 are shown in Figure 11b.The S-scheme composite photocatalyst with Bi-S bond was deliberately introduced at the interface between CdS quantum dots and Bi 2 WO 6 monolayer, which effectively promoted the separation and transmission of photogenerated carriers.
In the S-scheme charge transfer mechanism, carriers with weak redox ability are recombined, leaving photogenerated holes and electrons with strong redox ability, so that carriers complete spatial separation.With the deepening of research, the construction of double S-scheme heterojunctions with multichannel charge transfer has been proposed.Yogesh et al. synthesized a double S-schemes ZnIn 2 S 4 /BiOCl/FeVO 4 heterojunction.Compared with bare ZnIn 2 S 4 , BiOCl, and FeVO 4 , it shows significant photocatalytic efficiency for the degradation of RhB, reaching 98.06% in 30 min. [60]The results of free radical capture experiments and electron paramagnetic resonance (EPR) characterization confirm that RhB was degraded by the active substances of •OH and •O 2 À , and the synergistic effect of charge transfer in the double S-scheme is related to the improvement of photocatalytic activity and redox ability.In the future, the design of double S-scheme heterostructures for other materials can be carried out, and the internal charge transfer path can be clarified and applied to the analysis of reaction mechanisms in various fields.
Both the S-scheme and the Z-scheme heterojunction work on the same principle.Heterojunctions based on the S-scheme are made from two n-type semiconductors, which have appropriate Fermi levels.It is therefore rare for S-scheme heterojunctions to be distinguished from Z-scheme heterojunctions in most situations.Photogenerated carriers can be separated more efficiently and photocatalytic activity can be improved with both systems.

Photocatalytic Application of Metal Sulfur-Based Semiconductors
For the energy and environment crisis to be resolved, it is imperative that solar energy be developed and applied.Utilizing solar energy can be achieved through visible light response photocatalytic reactions.Photogenerated carriers react with O 2 and H 2 O to form active substances (•O 2 À and •OH) with strong oxidation.9][100] The application of photocatalytic behavior based on visible light includes: photocatalytic carbon dioxide reduction, photocatalytic water splitting, photocatalytic nitrogen fixation, photocatalytic pollutant degradation, and so on.Figure 12 presents a milestone schematic illustration of metal sulfide-based nanoarchitectures for energy and environmental applications.
In the photocatalytic effect, light energy is used to excite semiconductors.103] Many studies have shown that metal components or defects on Figure 11.a) Band structure of samples.Reproduced with permission. [95]Copyright 2022, Elsevier.b) Energy-band structures of CdS and Bi 2 WO 6 .Reproduced with permission. [96]Copyright 2021, Elsevier.
[106] The photogenerated electrons have strong reducibility and can react with the oxidant on the surface of the semiconductor, while the photogenerated holes have strong oxidizability and can react with the reducing agent.
Photogenerated carriers of the photocatalyst will be captured by the active sites on the semiconductor surface after separation and migration, and they will participate in the photocatalytic reaction with the help of the active sites.Metal sulfide-based photocatalysts have adjustable physicochemical properties, which may provide abundant catalytic active sites to promote light absorption, thereby promoting carrier migration and accelerating energy conversion between solar energy and chemical energy.The active site of the catalyst refers to the special position with high activity on the surface of the catalyst, which can adsorb the reactant molecules and undergo redox reactions with them, thereby accelerating the reaction process.In the photocatalytic reaction, only the active site can promote the catalytic reaction, and the surface of the remaining catalysts is inert, which is called the inert site.It mainly involves vacancy types, metal active sites, edge configurations, crystal surface engineering (Figure 13), and other types, [105,106] and is expected to improve energy efficiency and alleviate environmental issues (Table 2).The active sites of photocatalyst are mainly manipulated by optimizing the original active sites (crystal plane engineering, defect regulation) and introducing new active sites (cocatalyst, doping, heterojunction, morphology regulation).
The active sites in the catalyst are usually distributed at the surface and interface of the catalyst.In the process of photocatalytic reaction, these active sites are in constant contact with the reactant molecules and intermediates and react.Therefore, identifying the active sites accurately provides a favorable theoretical basis for exploring the photocatalytic mechanisms.In recent years, characterization techniques such as X-ray absorption spectroscopy (XAS), in situ diffuse reflectance Fourier transform infrared spectroscopy (DRIFTS), and in situ near-normal pressure X-ray photoelectron spectroscopy (NAP-XPS) can fully characterize active sites (coordination environment, interaction between sites and molecules, reaction intermediates and valence states, etc.). [104,107,108]XAS has high resolution and selectivity.It can determine the atomic composition (valence state of elements) and chemical structure (electronic structure, coordination environment) of the catalyst, and can detect the electronic structure and coordination environment information in the catalyst reaction process in real time.For instance, Benedikt et al. reported the use of in situ XAS to study the amorphous molybdenum sulfide (MoS x ) proton reduction catalyst, demonstrating that the terminal S 2 2À is involved in the hydrogen evolution reaction (HER). [107]In situ DRIFTS has high selectivity and high applicability, and can explore the distribution of active sites of the catalyst and the acidity and alkalinity of the material surface.Haandel et al. monitored the adsorption of NO by MoS 2 /Al 2 O 3 catalyst by infrared spectroscopy, and found that only a small number of edge sites may play the role of active sites in the catalytic reaction. [109]Yang et al. designed RuS 2 catalysts enriched with sulfur vacancies and identified the S vacancies in RuS 2Àx as the true active site for the catalytic reaction by in situ Raman spectroscopy. [108]n situ NAP-XPS is widely used in surface analysis and catalytic research.It can detect the element state under reaction conditions and give the element information and electronic structure information of the material surface.The CdS/WO 3 S-scheme heterojunction constructed by Liu et al. revealed by XPS that the larger spatial angle between the (100) crystallographic plane of CdS and the (111) crystallographic plane of WO 3 better preserves the highly active crystallographic planes and active sites. [110]

Photocatalytic Water Splitting
Water splitting is considered an attractive pathway for sustainable hydrogen production and involves two half-reactions, namely, HER [111] and oxygen evolution reaction. [112]In order to decompose water into H 2 and O 2 , the photocatalyst's oxidation and reduction potentials should be suitable.Water splitting is a nonspontaneous reaction that requires the outside world to provide a certain amount of energy for the reaction to occur.Photocatalysts trigger the reaction by converting solar energy into stored hydrogen energy according to the Gibbs energy we observe in the reaction. [113]From a thermodynamic perspective, photocatalytic water splitting requires that the VB of the semiconductor be more positive than the oxidation potential of water, that is, the VB of the photocatalyst needs to be higher than the energy level (at pH = 7, þ0.82 eV vs normal hydrogen electrode (NHE) of O 2 /H 2 O.The CB position is more negative than the proton reduction potential, that is, the CB of the photocatalyst needs to be lower than the energy level (at pH = 7, À0.41 eV vs NHE) of H þ /H 2 . [114]Photogenerated electrons reduce protons or water molecules adsorbed on the surface of the semiconductor to hydrogen, and photogenerated holes oxidize water molecules to generate oxygen.However, overall water splitting requires higher photon energy than 1.23 eV due to the large overpotential of the charge transfer process.Water splitting with photocatalysis produces hydrogen in some cases, but oxygen in others.As a result, photocatalytic water splitting is typically studied separately from the development of the overall water splitting system. [112]ased on the literature research, few studies have been conducted on the photocatalytic oxygen evolution of metal sulfide semiconductors, so we will focus on the progress of photocatalytic hydrogen evolution next.Many research works have been conducted on metal sulfide photocatalysts, which have shown good photocatalytic hydrogen evolution performance.For example, Zhang et al. prepared flower-like microsphere MoS 2 /ZnIn 2 S 4 composites assembled by nanosheets. [115]With visible light irradiation, it greatly improved the photocatalytic hydrogen production activity of ZnIn 2 S 4 .In the optimal 5% MoS 2 /ZnIn 2 S 4 system, the hydrogen production activity is as high as 3891.6 μmol g À1 h À1 (Figure 14a).The photocatalytic efficiency decreases after the third cycle (Figure 14b), which may be attributed to the photocorrosion of MoS 2 /ZnIn 2 S 4 composites during the prolonged reaction.As shown in Figure 14c, in the MoS 2 /ZnIn 2 S 4 heterostructure, the CB position of ZnIn 2 S 4 is more negative than the corresponding energy level of MoS 2 (À0.5 eV), and the photogenerated electrons can be transferred from ZnIn 2 S 4 to MoS 2 through the heterostructure interface.MoS 2 acts as a capture site for photogenerated electrons to inhibit the binding of photogenerated carriers.The electrons on MoS 2 will reduce H 2 O to H 2 .
Chong and co-workers constructed a hollow double-shell CdS@ZnIn 2 S 4 system.Pt and Co 3 O 4 nanoparticles were decorated on the inner and outer surfaces of CdS@ZnIn 2 S 4 to prepare a novel Pt/CdS@ZnIn 2 S 4 /Co 3 O 4 type II heterojunction photocatalyst to help hydrogen evolution (Figure 14d). [116]When CdS is integrated with ZnIn 2 S 4 /Co 3 O 4 , a built-in field is appeared inside the interface, which can promote the separation of electron-hole pairs, thereby improving the photocatalytic performance.In addition, the synergistic effect of spatially separated double catalysts is beneficial for migrating photogenerated charges in opposite direction and the inhibition of reverse reaction.As shown in Figure 14e, the hydrogen evolution rate reaches 8.53 mmol g À1 h À1 , and the H 2 O 2 yield reaches 5.26 mmol g À1 h À1 , which is 3.8 and 1.8 times that of CdS@ZnIn 2 S 4 /(PtþCo 3 O 4 ).It was found that the catalyst exhibited strong hydrogen production activity and excellent stability and durability (Figure 14f ).
Sun et al. designed and synthesized a particle-graded hollow tandem heterojunction photocatalyst (Figure 14g), which combines two heterojunctions of hollow black TiO 2 /MoS 2 nanosheets and MoS 2 nanosheets/CdS nanoparticles to improve the light utilization rate in a wide solar spectral range, thereby improving the photocatalytic efficiency. [117]To inhibit electron and hole recombination recombination, MoS 2 nanosheets act as cocatalysts and bridges to integrate hollow black TiO 2 and CdS into a particle tandem system (b-TiO 2 /MoS 2 /CdS).In contrast with the traditional single heterojunction structure, this structure is similar to the Z-scheme (Figure 14h).The obtained b-TiO 2 /MoS 2 /CdS microspheres with tandem heterojunction exhibit capacious spectral absorption, unified hierarchical hollow architecture, nanoparticle-modified nanosheet shell, and large surface area (Figure 14i).When utilized as a photocatalyst, it may produce 179 μmol h À1 hydrogen for every 20 mg of photocatalyst, which is almost 3 times that of the b-TiO 2 /MoS 2 heterojunction (57.2 μmol h À1 ).In comparison with MoS 2 /CdS, TiO 2 / MoS 2 /CdS heterojunction with CdS nanoparticles shows greater stability, which is due to the formation of tandem heterojunction and the strong UV absorption effect of black TiO 2 .
By comparing the photocatalytic hydrogen evolution rate of the above examples, the hollow tandem heterojunction photocatalyst b-TiO 2 /MoS 2 /CdS designed by Sun et al. showed better photocatalytic activity, which was attributed to the special morphology control and the design of Z-scheme heterojunction.These various strategies and regulations provide us with design ideas for optimizing the performance of photocatalysts (Table 3).

Photocatalytic Reduction of Nitrogen
Photocatalytic nitrogen fixation is a mild nitrogen fixation method that can replace the industrial Haber-Bosch reaction, which requires high temperature and high pressure to produce NH 3 .It is very difficult to directly dissociate the N≡N triple bond and initiate the hydrogenation reaction because the dissociation energy of the N≡N triple bond is as high as 941 kJ mol L À1 to generate NH 3 through the dissociation mechanism.Unlike the general photocatalytic reaction, photocatalytic nitrogen fixation needs to trigger the N 2 reduction mechanism (CB potential is higher than N 2 reduction potential).Under sunlight irradiation, the metal sulfide photocatalyst is excited by absorbing light energy to produce electron and hole pairs.These photogenerated carriers are active on the surface of the catalyst, react with N 2 adsorbed on the surface of the catalyst to activate N 2 , and convert it into NH 3 (Table 4).During N 2 activation and hydrogenation reduction, photocatalysis usually involves usually undergoes a proton-assisted multielectron transfer process.Two pathways may be involved in different photocatalytic systems, namely, remote binding and alternate binding.
Recently, Liu et al. in situ grown ultrathin MoS 2 nanosheets stacked nanoflowers on oxygen self-doped porous biochar (OPC) with precursor litchi peel (Figure 15a). [118]Due to the large specific surface area of OPC, N 2 can be adsorbed on the surface of OPC.Under simulated sunlight irradiation, in ultrathin MoS 2 ,  [115] Copyright 2018, Elsevier.d) Possible photocatalytic mechanism of Pt/CdS@ZnIn 2 S 4 /Co 3 O 4 .e) Photocatalytic hydrogen production rate of different catalysts.f ) Cyclic stability test of hydrogen production of Pt/CdS@ZnIn 2 S 4 /Co 3 O 4 sample.Reproduced with permission. [116]Copyright 2022, Elsevier.g) Synthesis of granular b-TiO 2 /MoS 2 composite catalyst/CdS tandem heterojunction and h) schematic diagram of solar-driven water splitting principle.i) Comparison of photocatalytic hydrogen production activity of different photocatalysts under visible light irradiation.Reproduced with permission. [117]Copyright 2018, Wiley-VCH.
electrons generated by photonics directly transfer to OPC surface, increasing photoinduced charge separation efficiency.Photofixation of N 2 is facilitated by the synergistic effects of MoS 2 and OPC.In comparison with the original MoS 2 , the composite has a higher ammonia production rate, which can deliver 156.4 μmol g À1 after 5 h (Figure 15b).The average ammonia yield of MoS 2 /OPC is 37.9 μmol g À1 h À1 , and it performs 1.6 times better than MoS 2 microspheres (Figure 15c).After three cycles of experiments, the hydrogen production is still above 24 μmol g À1 h À1 (Figure 15d).
Using an alkaline complexing agent, Dong and co-workers made the wurtzite and zinc blende phases coexist in a simplex Zn 1-x Cd x S nanocrystal, and prepared a nanocrystal with tunable bandgap and good stability. [119]Under visible light (λ ≥ 420 nm), the photocatalytic NH 3 production rate of 66.91 μmol g À1 h À1 and the apparent quantum efficiency (AQE) of 3.77% at 420 nm were achieved, which were about 17 times and 3 times higher than the original CdS and ZnS, respectively (Figure 15e).As a consequence, electron-hole pairs are separated and guided along their transmission paths by a homojunction between two crystal phases, preventing recombination.
Trying to integrate different crystal systems of photocatalysts or supplemented by other designs may be a good direction for regulating photocatalytic performance.Compared with the application of photocatalytic agents in other fields, the application of metal sulfide-based semiconductor photocatalysts in photocatalytic reduction of nitrogen is less.In future research, photocatalytic nitrogen fixation technology is expected to be further optimized to contribute to environmental purification.

Photocatalytic Reduction of Carbon Dioxide
Managing carbon balance and achieving carbon neutrality can be achieved by photocatalytic reduction of carbon dioxide to produce carbon-based fuels and high-value chemicals.As CO 2 is a linear molecule with a bond energy of about 750 kJ mol À1 , reducing it by photocatalysis requires a semiconductor capable of absorbing enough outside energy to destroy the C=O bond.In addition, photocatalytic CO 2 reduction is a multielectron reduction process with abundant reduction products (CH 4 , CH 3 OH, CO, etc.).These all mean that photocatalytic CO 2 reduction must face the dual demands on kinetics and thermodynamics.As shown in Figure 16a, most of the energy band positions of metal sulfide semiconductors satisfy the strong redox potential of CO 2 photoreduction (À1.90 eV vs NHE, pH = 7). [120]In recent years, they have shown great potential for photocatalytic CO 2 reduction.Under light irradiation, when the absorption energy of metal sulfide semiconductor photocatalysts is greater than its bandgap, the photogenerated carriers migrate, and the photogenerated electrons undergo a reduction reaction with CO 2 (carbon source) and H 2 O (hydrogen source) adsorbed on the surface of the photocatalyst, which can generate different reduction products such as CO, CH 4 , CH 3 OH, and HCOOH, [121] and the photogenerated holes undergo an oxidation reaction with H  [122] As shown in Figure 16b,c, the photocatalytic CO 2 production activity of CdS/Ni 9 S 8 /Al 2 O 3 system increases by 11 times (12.1 μmol h À1 ), which is much higher than that of CdS/Ni 9 S 8 (1.1 μmol h À1 ), and CdS/Ni 9 S 8 /Al 2 O 3 shows better photocatalytic stability than CdS/Ni 9 S 8 .
Using an optimized inorganic mixed reaction system, Meng and co-workers prepared Cd 2þ -modified colloidal ZnS photocatalyst for the reduction of CO 2 .The reduction product is mainly HCOOH, and the secondary product is CO. [123]The total yield of CO and HCOOH obtained in this study is 11.19 mmol h À1 g À1 (Figure 16e).In the presence of Cd 2þ grafted on ZnS nanocrystals, photogenerated electrons are effectively captured, and CO 2 molecules are activated and reduced (Figure 16d).Cd þ is formed by electron transfer from colloidal ZnS to Cd 2þ by single-photon absorption.Cd 2þ /Cd þ has a negative redox potential due to the instability of Cd þ , and Cd þ donates the first electron to CO 2 to form •CO 2 anion radical.In the reaction solution, the nucleophilic carbon atom of H 2 O gives a proton to the anion radical of •CO 2 at any time to form •HCO 2 .Finally, the second electron provided by Cd þ produced by Cd 2þ photoreduction will then reduce •HCO 2 to HCOO À (Figure 16f ).
Photocatalytic carbon dioxide reduction technology is a research hotspot in recent years.Although remarkable progress has been made (Table 5), it is still a great challenge for metal sulfide photocatalysts to increase light response range, reduce photogenerated carrier recombination, and accelerate the adsorption and activation of CO 2 .To obtain an appropriate energy band position, the design and construction of ternary heterojunctions may be a good strategy.

Photocatalytic Degradation of Pollutants
Environmental issue related to organic pollutants in water has gained widespread attention.Photocatalytic degradation of organic pollutants, as a new wastewater treatment technology with the advantages of environmental protection/energy saving/ sustainability, is considered to be the most effective solution.When the photocatalyst is irradiated by light energy larger or Photocatalysts have been studied extensively in recent years for their potential in reducing pollutants (Table 6).For example, Liu et al. prepared N-incorporated MoS 2 (N:MoS 2 ) nanoflowers with a surface area of 114.2 m 2 g À1 . [80]Active free radical capture experiment reveals that h þ , ⋅O 2 À , and ⋅OH act on the catalyst surface and generate oxidizing substances (Figure 17a).The Reproduced with permission. [118]Copyright 2020, Wiley-VCH.e) Photocatalytic N 2 fixation activity comparison between CdS, Zn 0.8 Cd 0.2 S, and ZnS.Reproduced with permission. [119]Copyright 2020, Elsevier.Reproduced with permission. [122]Copyright 2018, Elsevier.d) Schematic illustration of Cd 2þ cocatalyst for photogenerated electron separation and HCOOH selectivity of CO 2 reduction, e) HCOOH production corresponding to the cases in (a)-(d) within 8 h.f ) Schematic illustration of the possible HCOO À formation mechanism over Cd 2þ modified colloidal ZnS.Reproduced with permission. [123]Copyright 2017, Elsevier.17b): Using the optimized 20 mg N:MoS 2 nanoflower catalyst, 50 mL 30 mg L À1 RhB can be degraded entirely within 70 min, showing a first-order kinetics constant (k) of 0.06928 min À1 , which is 26.4 times more efficient than the pristine nanosheet (Figure 17c,d), and after four consecutive experiments, N:MoS 2 nanoflowers show good recycling and structural stability, and they are not significantly reduced in photocatalytic activity (Figure 17e).
According to Ouyang et al. it was demonstrated that the direct Z-scheme photocatalyst of ZnIn 2 S 4 @MoO 3 is capable of degrading tetracycline hydrochloride (TC-HCl) in the visible light range with good catalytic activity, and in comparison with original ZnIn 2 S 4 and MoO 3 , its reaction rate is 7.8 times and 25.8 times higher, respectively (Figure 17f ). [92]When the amount of indium zinc sulfide precursor is 3 times (Z@M-3), the degradation rate is the highest, reaching 94.5% (Figure 17g).As a result of the Z-scheme charge transfer mechanism within the heterojunction, photoinduced charge transfers can be mediated, the charge transfer can be promoted, and photogenerated electrons or holes can maintain their high redox abilities; the ZnIn 2 S 4 @MoO 3 composite catalyst shows significantly improved activity.
According to Tan et al. the nature of the crystal plane exposed to the surface is crucial to the photocatalytic activity of the photocatalyst. [124]The catalytic performance of ZnIn 2 S 4 microspheres doped with Sm 3þ on the exposed negative crystal plane for the degradation of MO and RhB under visible light was investigated.Upon visible light irradiation for 90 min, ZnIn 2 S 4 microspheres containing 2 mol% Sm 3þ can degrade RhB completely, but MO degradation is poor, and the degradation efficiency is not much different from pure ZnIn 2 S 4 microspheres (Figure 17h,i).This is due to the fact that hexagonal ZnIn 2 S 4 has a layered structure composed of S-Zn-S-In-S-In-S-In-S atomic layers along the {0001} direction.Negative S ions are exposed on the surface of {0001} plane (Figure 17j).Anionic dye MO molecules have a small interaction with the negatively charged S surface,  17k).
Therefore, defect design or adjustment of the crystal plane exposes more active sites, and a thorough understanding of the mechanism of this mystery still needs more research work.In addition, combined with the research results of others, it is not difficult to find that most photocatalysts can effectively degrade dyes, but when encountering organic compounds with complex structures such as antibiotics, the degradation performance is reduced.Consequently, it is of great significance for human  Reproduced with permission. [80]Copyright 2016, IOP Publishing Ltd.f ) Apparent reaction rate constants for the degradation of TC-HCl.g) Photocatalytic degradation of TC-HCl curves.Reproduced with permission. [92]Copyright 2021, Springer Nature.h) Photodegradation curves of undoped and Sm 3þ -doped ZnIn 2 S 4 for RhB and i) MO under visible light.j) The {0001}-faceted crystal structure of ZnIn 2 S 4 and its RhB degradation mechanism.k) Degradation mechanism of RhB on Sm-doped ZnIn 2 S 4 under visible light.Reproduced with permission. [124]Copyright 2014, Elsevier.
survival and development to find or design materials with more photocatalytic activity to degrade antibiotic drugs.

Summary and Outlook
Semiconductor materials can effectively convert visible light into other forms of energy, which is a research hotspot that has attracted much attention.Metal sulfides are reduced semiconductor photocatalysts with narrow bandgap and negative CB potential, which has aroused significant concern in developing visible light-responsive photocatalysts.First, this article discusses diverse synthetic approaches used to create metal sulfide semiconductor photocatalysts, such as solid-phase synthesis, liquid-phase preparation, gas-phase fabrication, and other synthesis methods.Second, we present strategies for optimizing the catalytic performance of metal sulfide-based photocatalysts by manipulating morphology, doping engineering, and constructing heterojunctions.Solar energy optimization and photogenerated carrier separation are responsible for the excellent photocatalytic performance.Photogenerated electrons and holes can be separated very well using Z-scheme and S-scheme heterojunctions, which are more preferred than other approaches.Therefore, good energy band match and reasonable heterostructures are essential requirements for the metal sulfide-based semiconductor photocatalysts.Next, the characterization methods for accurately identifying active sites are summarized.In-depth study of active sites will help to improve the selectivity, activity, and lifetime of photocatalysts in the future and better understand the important reaction mechanisms of photocatalytic reactions.Finally, the applications of various metal sulfide-based semiconductors under visible light irradiation are methodically expounded, including the ability to decompose water into oxygen and hydrogen, which provides new avenues for energy production development; degradation of organic pollutants and improvement of the environment are possible with metal sulfides; it can effectively improve the reduction of carbon dioxide and nitrogen.
In the areas of energy development and environmental governance, metal sulfide-based semiconductor photocatalysts with appropriate narrow bandgaps are being developed.Despite some progress, metal sulfide semiconductor photocatalysts still face many obstacles and challenges. 1) Design and create a richer form of metal sulfide-based semiconductor material photocatalyst structure.High-quality metal sulfide-based semiconductor materials can be obtained by improving existing preparation methods, exploring new technologies, combining morphology (such as nanosheets, nanorods, microspheres, hierarchical structures, etc.), combining doping strategies (heterovalent ion doping, anion-cation mixed doping, etc.), and constructing heterojunctions.2) Improving the stability of metal sulfide-based photocatalysts is still the focus of research.Further development of metal sulfide photocatalysts with improved stability and exploration of the activity and deactivation mechanisms of these photocatalysts using advanced in situ characterization methods.3) Utilizing cutting-edge defect engineering to manipulate the atomic-scale defects (point, line, surface defects) of such materials by taking advantage of the easy lattice reconstruction of inorganic compounds, identifying the distribution of active sites, and quantitatively analyzing the correlation between active sites and photocatalytic properties.4) It is still of great significance to pave the way of existing metal sulfide-based photocatalysts from the laboratory to industrialization, including the establishment of industrialgrade photocatalyst production lines, the manufacturing process of ton-level photocatalysts, and the catalytic performance of photocatalysts after mass production.

Figure 1 .
Figure 1.A summary of the synthetic methods for the synthesis of metal sulfur-based compounds.

Figure 2 .
Figure 2. a) Schematic diagram of high energy ball milling reaction.b) transmission electron microscope (TEM) image of ZnS/C nanocomposite.Reproduced with permission.[31]Copyright 2018, Elsevier.c) Schematic illustration of high temperature solid-phase reaction.d) scanning electron microscope (SEM) images of ZnS/ZnO at different reaction times.Reproduced with permission.[33]Copyright 2023, Elsevier.

Figure 6 .
Figure 6.Schematic illustrations of band position changes caused by the doping of metal elements (left) and nonmetal elements (right).

Figure 8 .
Figure 8. a) Schematic diagram of the energy level configuration of g-C 3 N 4 /CdS.b,c) Photocatalytic mechanism of hydrogen production on a CdS nanowire with a thin and unclosed g-C 3 N 4 shell.Reproduced with permission.[84]Copyright 2013, American Chemical Society.d) Photocatalytic mechanism for degrading RhB by In 2 S 3 /Bi 2 MoO 6 .Reproduced with permission.[85]Copyright 2020, Elsevier.e) Schematic illustration of the band edge positions for both La 2 Ti 2 O 7 and In 2 S 3 .f,g) Proposed band structure and photocatalytic reactions for La 2 Ti 2 O 7 /In 2 S 3 nanosheet heterojunctions.Reproduced with permission.[86]Copyright 2019, Elsevier.

Figure 15 .
Figure 15.a) Formation mechanism of MoS 2 /OPC.b) NH 3 yield rates after consecutive photocatalysis for 5 h.c) The average of NH 3 yield rate of the three samples.d) NH 3 yield rate of MoS 2 /OPC during three consecutive cycles.Reproduced with permission.[118]Copyright 2020, Wiley-VCH.e) Photocatalytic N 2 fixation activity comparison between CdS, Zn 0.8 Cd 0.2 S, and ZnS.Reproduced with permission.[119]Copyright 2020, Elsevier.

Figure 16 .
Figure 16.a) Band positions of representative metal sulfide semiconductors and the redox potential of CO 2 for NHE at pH = 7. b) Precipitation rate of CO in the photoreduction of CO 2 and H 2 O on Al 2 O 3 driven by visible light.c) Stability in visible-light-driven CO 2 photoreduction.Reproduced with permission.[122]Copyright 2018, Elsevier.d) Schematic illustration of Cd 2þ cocatalyst for photogenerated electron separation and HCOOH selectivity of CO 2 reduction, e) HCOOH production corresponding to the cases in (a)-(d) within 8 h.f ) Schematic illustration of the possible HCOO À formation mechanism over Cd 2þ modified colloidal ZnS.Reproduced with permission.[123]Copyright 2017, Elsevier.

Figure 17 .
Figure 17.a) Free radical capture experiments and b) RhB degradation schematics of N-doped MoS 2 nanoflowers.c) Photocatalytic degradation curves of RhB by different photocatalysts under visible light irradiation.d) Variation curves of ln (C 0 /C) of N:MoS 2 nanoflowers and MoS 2 nanosheets with irradiation time.e) Cyclic experiment of RhB degradation by N:MoS 2 nanoflowers.Reproduced with permission.[80]Copyright 2016, IOP Publishing Ltd.f ) Apparent reaction rate constants for the degradation of TC-HCl.g) Photocatalytic degradation of TC-HCl curves.Reproduced with permission.[92]Copyright 2021, Springer Nature.h) Photodegradation curves of undoped and Sm 3þ -doped ZnIn 2 S 4 for RhB and i) MO under visible light.j) The {0001}-faceted crystal structure of ZnIn 2 S 4 and its RhB degradation mechanism.k) Degradation mechanism of RhB on Sm-doped ZnIn 2 S 4 under visible light.Reproduced with permission.[124]Copyright 2014, Elsevier.

Table 2 .
Comparison of several typical active sites and photocatalytic properties of metal sulfide-based photocatalysts in photocatalytic applications.

Table 3 .
Summary of the photocatalytic water splitting activity of metal sulfide-based compound photocatalysts.
2 O to produce O 2 .Li et al. synthesized a CdS/Ni 9 S 8 /Al 2 O 3 ternary photocatalyst by immobilizing CdS nanoparticles on Ni 9 S 8 /Al 2 O 3 substrate, which significantly improved the CO 2 conversion efficiency.

Table 4 .
Overview of photocatalytic N 2 fixation performance of metal sulfide-based compound photocatalysts.
equal to its bandgap, the photogenerated carriers are separated and migrated to the active sites on the surface of the photocatalyst under the action of IEF.Photogenerated electrons have strong reducibility and are easily captured by dissolved oxygen adsorbed on the semiconductor, and oxygen captures electrons to form ⋅O 2 À ; the photogenerated holes have strong oxidizing properties, which can oxidize H 2 O and OH À adsorbed on the surface of the catalyst into highly active •OH radicals.The active •OH radicals oxidize organic pollutants such as dyes and antibiotics into small molecules of CO 2 and H 2 O.

Table 5 .
Overview of the photocatalytic CO 2 reduction activity of metal sulfide-based compound photocatalysts.À1 g À1 ] compared to the positively charged RhB molecules.Sm-doped ZnIn 2 S 4 sample can trap photoinduced electrons and prevent electron-hole pair recombination by substituting Sm 3þ for In 3þ in the lattice.Sm 3þ /Sm 2þ doping levels are just below the ZnIn 2 S 4 CB and electrons can be excited into the doping levels under visible light.Therefore, Sm 3þ can easily capture electrons from CB, and the doped energy levels and captured electrons of Sm 3þ /Sm 2þ can react with O 2 to yield H 2 O 2 /HO• radicals (Figure

Table 6 .
Overview of the photocatalytic degradation activity of metal sulfide-based compound photocatalysts.