Layered Trichalcogenidophosphate: A New Catalyst Family for Water Splitting

With the rapidly increasing demand for energy and extensive use of fossil fuels, there is an urgent need to find alternative sources of clean energy to solve the ever-growing environmental problems. Hydrogen, the simplest and most abundant element in the universe, has a high energy density and is sustainable and eco-friendly. The combustion of hydrogen with air/oxygen offers high energy and gives off water as the end product. Hence, hydrogen is considered one of the most promising candidates for fulfilling future energy demands [1, 2]. It has been long recognized that water can be decomposed under solar excitation in plants. Electrical power can also be used to facilitate water splitting into gaseous hydrogen and oxygen, which can be stored for further use. However, it is still a huge challenge to split water on the industrial scale as efficiently as in plants. Moreover, the state-of-the-art electrocatalytic techniques for water splitting still rely heavily on noble metal catalysts such as Pt for the hydrogen evolution reaction (HER) and IrO₂ for the oxygen evolution reaction (OER). Thus, the generation of hydrogen and oxygen is a high-cost operation, making the use of hydrogen and oxygen as commercial fuels challenging. Therefore, it is highly desirable to find earth-abundant, cost-effective, and highly active catalysts for water splitting [3].

As requirements of effective photocatalytic water splitting, the catalysts should possess a suitable bandgap and carrier mobility [4]. The photo-generated excitons (commonly known as electron–hole pairs) must migrate separately to the surface of the catalyst and be consumed by corresponding redox reactions. In order to prevent recombination of the excited electron and hole, the redox reactions should proceed within the lifetime of these species [4, 5, 6, 7]. In this regard, band structure engineering and charge carrier concentration tuning are highly important. Further, as key factors, the catalyst must enable fast charge transfer kinetics and lower the energy for chemisorption of various intermediate species on the catalyst surface to improve the electrochemical water splitting efficiency [8, 9, 10, 11]. It has been demonstrated that the chemisorption energies of intermediate species on the catalyst surface can be adjusted by varying the chemical composition of the electrocatalysts [10]. For instance, phosphorus-doped pyrite-type cobalt disulfide exhibits excellent hydrogen evolution activity because of effective reduction in the energy for adsorption of atomic hydrogen on the adjacent cobalt sites doped with phosphorus atoms [11, 12, 13]. Therefore, in order to meet the above-mentioned requirements, variation of the composition of the catalyst should be effective for adjusting the bandgap and electronic structures, charge carrier concentration and mobility modulation, and regulation of the chemisorption energy in semiconductor-based catalytic systems.

Layered metal trichalcogenidophosphates, known as metal phosphorus trichalcogenides, have attracted increasing attention in recent years. Since the first discovery of metal chalcogenidophosphate by Friedel more than a century ago [14], a large family of these layered compounds has been synthesized and studied. The general formula for these compounds can be expressed as M^IIPQ₃ or M _0.5 ^I M _0.5 ^III PQ₃, where M^II represents bivalent metals such as Mg, V, Mn, Fe, Co, Ni, Zn, Cd, Sn, and Hg; M^I represents metal ions such as Cu and Ag; and M^III can be Cr, V, Al, Ga, In, Bi, Sc, Er, or Tm; Q is a chalcogen, i.e., S or Se (to be more concise, the abbreviation MPQ₃ is used hereinafter) [14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47]. The diversity of metal, phosphorus, and chalcogen atoms in MPQ₃ materials offers vast possibilities for achieving the desired physical, chemical, optical, and electronic properties. For instance, the presence of sulfur and phosphorous in metal trithiophosphate compounds exerts a synergistic effect on the surface electronic structure of the central metal atoms [48, 49]. Moreover, the asymmetric M^I and M^III cations possess different metal–chalcogen bond distances, resulting in differences in the valence band maximum (VBM) and conduction band minimum (CBM) orbital distribution [49]. Moreover, due to the layered structure of MPQ₃ materials, they can be easily prepared as two-dimensional nanostructures having a large surface area and numerous exposed active sites. Due to these merits, MPQ₃ materials are widely considered good candidates as high-performance photo- or electrocatalytic water splitting catalysts.

Although there have been several reviews involving layered MPQ₃ in recent years, there is still no specific summary on layered MPQ₃ materials. In addition, it is important to summarize the progress on the MPQ₃ materials used for photo- and electrocatalytic water splitting due to the recent rapid development. In this regard, we aim to provide a brief summary of recent work and offer an overview of this promising field. In this minireview, we provide a brief introduction of the structure of MPQ₃ materials and subsequently discuss the methods of synthesis, highlighting the advantages and limitations. In the following sections, recent developments of MPQ₃ materials for photo- and electrocatalytic water splitting are briefly summarized. The roles of MPQ₃ materials in different reaction systems are also discussed. Finally, the challenges and prospects of MPQ₃ materials are proposed on the basis of their current developments.

Generally, trithio- and triseleno-phosphate compounds are considered to have a layered crystal structure similar to that of transition metal disulfides (TMDs, e.g., MoS₂). Figure 1 shows the typical crystal structures of MoS₂, M^IIPQ₃, and Ag_0.5M _0.5 ^III PQ₃. In contrast to the structure of MoS₂, P–P pairs substitute one third of the metal atoms within the MPQ₃ layer. Therefore, each MQ₆ octahedron is surrounded by three P₂Q₆ groups. The metal layer in MPQ₃ is encapsulated by both chalcogens and phosphorus atoms. In the case of Ag_0.5M _0.5 ^III PQ₃ particularly, the asymmetric Ag^I and M^III cations possess different metal–chalcogen binding distances, resulting in a distorted P₂Q₆ polyhedron (Fig. 1c, Ag_0.5M _0.5 ^III PQ₃) [49].

Open image in new window

Conventionally, high-quality MPQ₃ single crystals are synthesized by the chemical vapor transport (CVT) method, in which pure elemental powders are used as the starting materials and a small amount of iodine is used as the transport agent. These starting materials are vacuum-sealed in a quartz tube, which is then placed in a horizontal tube furnace with a temperature gradient. During the CVT process, the transport agent carries the starting materials from the hot end to the cold end of the tube for crystal growth. The crystal growth temperature is usually higher than 600 °C and may vary depending on the metal species. The growth period ranges from several days to even 2 months [24]. Finally, in order to obtain the nanosheets (i.e., single- or few-layered MPQ₃ products), exfoliation techniques such as micromechanical exfoliation or liquid exfoliation are carried out [43, 50, 51]. Limited by the high reaction temperature, long growth period, and possible explosion hazards, the CVT method is unsuitable for large-scale production of MPQ₃ materials.

Compared with the CVT method, chemical vapor deposition (CVD) is more suitable for the large-scale growth of MPQ₃. In a modified CVD method for the synthesis of NiPS₃ [3], elemental sulfur and phosphorus are evaporated in the two-zoned tube furnace under Ar flow; Ni(OH)₂ nanosheets grown on various substrates as a metal source are simultaneously heated in another zone. The reaction time for this CVD process is only 1 h, which is greatly reduced when compared with the CVT method. Due to the non-exclusive reaction environment and the reduced reaction time, the CVD process is considered a safer and faster method for producing MPQ₃ nanosheets. However, the requirement for a highly pure inert atmosphere and complex reaction conditions may limit scalable production.

Recently, a method termed solid-state transformation (SST) was developed for the large-scale production of MPQ₃ materials [4]. In a typical SST process, metal hydroxides are chosen as starting materials, and elemental sulfur and phosphorus are directly ground with the hydroxides. The mixture is vacuum-sealed in a Pyrex or quartz tube and heated at a certain temperature. After removing the excess sulfur and phosphorus, the MPQ₃ materials can be obtained. Depending on the compositions and morphologies of the starting metal hydroxides, the products can be either nanosheets or other nanostructures. The reaction temperature of the SST method may be even lower than that of the CVT and CVD methods. With such a low reaction temperature and no requirement for the two-zone furnace, the SST method is regarded as a promising technique for mass production of MPQ₃.

The bandgap structures of many single-layered group II trichalcogenidophosphates (M^IIPQ₃, M^II = Mg, Zn, Cd) and group III silver trichalcogenidophosphates (Ag_0.5M _0.5 ^III PQ₃, M^III = Sc, In) have been studied based on density functional theory (DFT) calculations [49]. As shown in Fig. 2a, Liu and Peng aligned the CBM and VBM of these compounds with respect to the water redox potential levels and evaluated their redox abilities. In consideration of the stability of single-layered MPQ₃ and the changes in the redox potential based on the pH, Zhou and co-workers also studied the band edges and the optical properties of MPS₃ (M = Mn, Fe, Ni, Zn, Cd) and MPSe₃ (M = Mn, Fe) using theoretical calculations [55]. Using the theoretically calculated optical absorption coefficient (α) of a series of single-layer MPQ₃ species, MnPSe₃, FePQ₃, and NiPS₃ were found to exhibit optical absorption in the visible spectral range (1.59–3.26 eV) [55]. A similar tendency was also experimentally observed by Kloc and Xiong [50]. Notably, among the studied compounds, FePSe₃ was found to exhibit the narrowest bandgap based on both calculated and experimental data. However, FePSe₃ was not suitable for water splitting at pH = 7 as the CBM is lower than the reduction potential of H⁺/H₂, whereas MnPSe₃ showed strong adsorption in the visible spectral region. Further, the calculated carrier mobility of single-layered MnPSe₃ is much higher than that of many other 2D materials (electron mobility: 625.9 cm² V⁻¹ S⁻¹; hole mobility: 34.7 cm² V⁻¹ S⁻¹). The huge difference between the mobility of these two carriers suggests effective separation of the photo-generated electron–hole pairs during the photocatalytic process. This makes MnPSe₃ a promising candidate for high-efficiency photocatalytic water splitting.

Open image in new window

Apart from the individual materials, the combination of photoactive materials with other co-catalysts to form an integrated system is another important field in photocatalytic research. Therefore, it is necessary to study the electronic structure of the heterojunctions in these integrated systems. Mi and co-workers recently explored the electronic structure of MnPSe₃-based 2D van der Waals heterostructures using DFT calculations. In the case of the MnPSe₃/MoS₂ heterostructure [56], different stacking patterns of single-layered MnPSe₃ and MoS₂ were studied, as presented in Fig. 3a. In some stacking patterns, spin splitting at the VBM of the MnPSe₃ layer was observed due to hybridization of the d orbital in Mn, thus enhancing the electron mobility (Fig. 3b). In addition, MnPSe₃/MoS₂ is a type II heterostructure in which the top part of the valence band is mainly contributed by MnPSe₃, while the bottom part of the conduction band is mainly contributed by MoS₂. Typically, the type I heterostructure has a symmetrical offset of potential barriers for the electrons and holes, where direct exciton transition occurs at the heterointerface. In type II heterostructures, the electrons and holes are localized on different sides of the heterointerface, which results in an indirect exciton transition [57, 58, 59]. The type II band alignment results in separation of the photo-generated electrons and holes, thus enhancing the photocatalytic efficiency. They also calculated the electronic structure of MnPSe₃/CrSiTe₃, where MnPSe₃ and CrSiTe₃ possess similar crystal structures and have a low lattice mismatch (about 4.7%) [60]. In this work, Mi et al. studied the effects of strain and an electric field on the heterostructures. When the heterostructure was formed under tensile strain, the band-edge position shifted from type I to type II, accompanied by a transition from an indirect bandgap to direct bandgap. When a compressive strain was applied, the heterostructure changed from semiconducting to conducting. At the same time, the band alignment could be tuned to type I or type II by applying a suitable electric field. The two mentioned studies demonstrated the possibility of band structure modulation of the MnPSe₃-based van der Waals heterostructure, suggesting the potential applicability of the novel heterostructures in photo-electronics.

Open image in new window

Although MnPSe₃ has gained extensive attention and shows great potential in the field of photocatalytic water splitting, experimental studies are still focused on the fundamental physical properties [50, 61]. Recently, He and co-workers turned their attention to Fe- and Ni-based compounds instead of Mn-based trichalcogenidophosphates [62, 63]. They conducted an experimental study of few-layered nickel trithiophosphate (NiPS₃) prepared by a modified CVD process for photocatalytic hydrogen evolution [62]. The resultant NiPS₃ nanosheets had a lateral size larger than 15 μm and thickness of less than 3.5 nm. The as-synthesized ultrathin NiPS₃ nanosheets could be used for photo-hydrogen evolution in neutral and pure water under xenon light or simulated AM1.5G solar illumination with a constant evolution rate of ~ 26.42 μmol g⁻¹ h⁻¹. However, the intrinsic driving force was not energetically sufficient for oxygen evolution due to the misaligned valance band of NiPS₃ relative to the water oxidation potential. Therefore, the NiPS₃ nanosheets suffered from a disadvantage similar to that of many other single-component photocatalysts, wherein the photoactivity is degraded by photo-anodic corrosion. This is because during the photocatalytic process, the photo-generated electrons are consumed by H₂ generation, while the photo-generated holes accumulate due to the misaligned valance band of NiPS₃ relative to the water oxidation potential. The highly oxidizing photo-generated holes then react with NiPS₃ itself [53]. Consequently, this study experimentally confirms the photocatalytic activity of NiPS₃. Nevertheless, the performance of NiPS₃ is much poorer than that of the commercially available materials, and this material is still far from practical application [7, 53]. Further study toward the design and construction of novel MPQ₃-based photocatalysts is still needed.

The photocatalytic hydrogen evolution performance of FePS₃ compounds was also recently studied [63]. He and co-workers developed a two-step hydrazine intercalation and exfoliation process for the synthesis of single-layered FePS₃ quantum sheets from bulk FePS₃ crystals. The resulting quantum sheets possessed a lateral size of 4–8 nm and thickness of less than 2 nm (Fig. 4a). As discussed above, the bandgap is closely related to the photocatalytic activity. In the case of the FePS₃ quantum sheets, the decreased size from bulk to quantum sheets increased the bandgap (Fig. 4b). Therefore, the hydrogen production efficiency of the FePS₃ quantum sheets was three times higher (290 μmol g⁻¹ h⁻¹) than that of the bulk FePS₃ (Fig. 4c). This value is around 11-fold higher than that of the NiPS₃ nanosheets [62], but still about an order of magnitude lower than that of the MoS₂-based materials [64] and Ni/Cu-modified titania [65]. Furthermore, measurement of the hydrogen evolution over 40 h indicated that these FePS₃ quantum sheets were more stable than the NiPS₃ nanosheets (Fig. 4d). Given that the reduced size of the quantum sheets leads to the exposure of numerous boundaries and active sites, the authors attributed the enhanced performance to the effective separation of photo-excited electrons and holes. When FePS₃ and NiPS₃ are compared with other MPQ₃ compounds, the relatively narrower bandgap of FePS₃ and NiPS₃ is thought to be the main contributor to their visible light photocatalytic activity. However, the narrow bandgap also results in insufficient exciton energy. Therefore, dopant metals such as Mn, Zn, and Cd that can broaden the bandgap might be helpful for bandgap engineering and heterostructure construction.

Open image in new window

Photocatalytic water splitting relies strongly on the absorption of solar energy, making this technique generally inefficient. In contrast, electrochemical catalysis, as one of the fundamental processes in electrochemistry, offers a more feasible and practical option [1]. In the past several decades, researchers have paid more attention to non-noble metal-based electrocatalysts due to their low cost. As a kind of semiconductor with tunable and suitable physiochemical properties, layered MPQ₃ compounds were also evaluated for electrocatalytic hydrogen evolution. A general mechanism for the electrochemical HER process might include an adsorption step and desorption steps as follows [66, 67, 68]:

Recently, Sampath and co-workers reported the HER performance of liquid-exfoliated NiPS₃ nanosheets over a wide pH range of 1–14 and in a neutral 3.5 wt% NaCl solution that simulates seawater [67]. The bulk NiPS₃ crystals were firstly synthesized through the CVT method. After exfoliation in the liquid phase, NiPS₃ nanosheets with a lateral size of 200–400 nm and thickness of 0.65–0.7 nm were obtained. The exfoliated NiPS₃ nanosheets showed improved electrochemical HER performance under all the pH conditions. The overpotential of the exfoliated NiPS₃ nanosheets at 10 mA cm⁻² in pH = 1, 14, and neutral 3.5 wt% NaCl solution was 297, 398, and 816 mV, respectively. The authors also studied the HER activity through DFT calculation. By evaluating the free energy of hydrogen adsorption (ΔG_H), they concluded that the phosphorus atom is the preferred site for H adsorption, consistent with the inference from metal phosphides [69, 70]. Later, Sampath’s group reported a similar study on the HER performance of exfoliated FePS₃ nanosheets [71]. The FePS₃ nanosheets demonstrated better HER performance than the Ni-based analogs, where the overpotential at 10 mA cm⁻² in pH = 1, 14, and neutral 3.5 wt% NaCl solution was 211 ± 3, 337 ± 4, and 637 ± 4 mV, respectively. When combined with reduced graphene oxide (rGO), the HER performance of both compounds could be further improved. The authors ascribed this enhancement to the improved conductivity derived from rGO.

Pumera et al. [68] also reported a systematic study of the electrocatalytic HER performance under acidic conditions for a series of MPS₃ compounds, including MnPS₃, FePS₃, CoPS₃, NiPS₃, ZnPS₃, CdPS₃, and SnPS₃. The overpotentials of NiPS₃ and CoPS₃ were 530 and 590 mV at a current density of 10 mA cm⁻², respectively. Here, the overpotential of FePS₃ (860 mV) was found to be higher than that of its Ni and Co analogs. The overpotentials of the other MPS₃ compounds were all higher than that of the bare glassy carbon electrode. According to their results, it is interesting that bulk NiPS₃ had a lower overpotential than bulk FePS₃, whereas Sampath et al. reported that exfoliated NiPS₃ nanosheets possess a higher overpotential than exfoliated FePS₃ nanosheets. Therefore, further study still needs to be conducted to explore the intrinsic properties of these materials.

Beyond the single-metal MPS₃ compounds, a series of Ni/Co bimetal trithiophosphate nanosheets (Ni_1−xCo_xPS₃, x = 0, 0.03, 0.05, 0.07, and 0.09) were prepared, and their HER performance under alkaline conditions (pH = 14) was evaluated [72]. The few-layered Ni_1−xCo_xPS₃ nanosheets were obtained by liquid exfoliation. Importantly, after Co doping, the HER performance of these Ni_1−xCo_xPS₃ nanosheets improved dramatically compared with that of the NiPS₃ nanosheets. Specifically, Ni_0.95Co_0.05PS₃ exhibited superior HER activity, with an overpotential of 71 mV at a current density of 10 mA cm⁻² and a Tafel slope of 77 mV dec⁻¹ (Fig. 5a, b). This performance was much better than that of the NiPS₃ and FePS₃ nanosheets supported by conductive rGO [67, 71]. Electrical conductivity measurements showed that Co doping can dramatically improve the conductivity of NiPS₃ (by about three orders of magnitude), suggesting changes in the electronic structure of NiPS₃ after Co doping.

Open image in new window

It is well known that Co is quite expensive compared with other transition metals like Mn, Fe, and Ni. Moreover, compared with Co, Fe is earth abundant and more common in high-efficiency hydrogenase. The incorporation of Fe with Ni might also be reasonable for modulating the HER activity of NiPS₃-based compounds. Therefore, Jin et al. [73] prepared a series of Ni/Fe bimetal trithiophosphate nanosheets (Ni_1−xFe_xPS₃ samples, x = 0, 0.05, 0.1, 0.15, and 1) as electrocatalysts for the HER. Through DFT calculation of ΔG_H, the authors proposed that H atom adsorption on edge sites is more favorable than on the basal surface. In addition, the ΔG_H for the S sites in NiPS₃ is negative, while it is positive for the S and P sites of FePS₃, implying the existence of optimal intermediates with better HER activity. The results confirmed the speculation that Fe-doped NiPS₃ can exhibit electrocatalytic HER performance comparable that of the Co-doped counterpart. For the optimized Ni_0.9Fe_0.1PS₃ compound, the overpotential at 10 mA cm⁻² was 72 mV and the Tafel slope could reach 73 mV dec⁻¹ (Fig. 5c, d).

In recent work, Pumera et al. [68] compared the OER activity of a series of bulk MPS₃ compounds in 1 m KOH solution. The results showed that most of the MPS₃ compounds exhibited low OER activity, whereas only CoPS₃ showed higher OER activity than the other analogs. However, the results are dubious because CoPS₃ can afford a current density of 10 mA cm⁻² at a potential of 0.84 V versus RHE, which is less than 1.23 V. Recently, our group evaluated the OER performance of bulk and liquid-exfoliated NiPS₃ nanosheets under alkaline conditions [51]. The exfoliated NiPS₃ nanosheets provided an overpotential of 301 mV@10 mA cm⁻², which is about 140 mV lower than that of bulk NiPS₃. Comparison of the electrochemically active surface area (ECSA) of these two materials indicates that the ECSA of the exfoliated NiPS₃ nanosheets is about 3.4-fold larger than that of bulk NiPS₃. Further, the current density of the exfoliated NiPS₃ nanosheets at 1.55 V is more than fivefold higher than that of bulk NiPS₃. The nonlinear increase in the current density, along with the ECSA, indicates that the large ECSA of the exfoliated NiPS₃ nanosheets is not the only reason for the improved OER activity.

A more detailed OER study of exfoliated NiPS₃ nanosheets was reported by Schuhmann et al. [48]. As mentioned above, the free radical species generated during the OER process might react with the catalyst. As such, surface oxidation cannot be avoided for MPS₃ materials. By employing a combination of scanning electrochemical microscopy (SECM), in situ Raman spectroscopy, SEM, and XPS measurements, Schuhmann et al. demonstrated that during the OER process, the NiPS₃ surface is oxidized, resulting in the formation of a NiPS₃@amorphous NiOOH core–shell heterostructure. According to their studies, the surface oxidation process can be divided into two steps. In the first step, the NiPS₃ surface is oxidized to Ni(OH)₂ when immersed in KOH solution. With increasing applied voltage, Ni²⁺ is further oxidized to Ni³⁺, accompanied by the transformation of Ni(OH)₂ to NiOOH before oxygen evolution. This detailed oxidation mechanism was not provided for previously reported NiS₂/NiS and NiCoP/C nanoboxes and other MPQ₃ compounds [51, 68, 76, 77]. Notably, the core–shell heterostructure can provide a current density of 10 mA cm⁻² with an overpotential of 350 mV, which is lower than that of noble metal oxide catalysts [78, 79]. Furthermore, DFT calculation indicates that the metallic character of the NiPS₃ nanosheets can provide an electron conduction pathway for efficient transport to surface NiOOH species. The NiOOH species at the surface with a high density of accessible active edges and defect sites can act as oxygen evolution centers. The verification of the relationship between this uniquely formed in situ core–shell heterostructure and the outstanding OER activity provides a useful guideline for understanding the OER mechanism in MPS₃ materials.

Studies on the OER performance of Ni/Fe bimetal trithiophosphates have also been documented [73, 80]. As reported by Jin and co-workers [73], Ni/Fe bimetal trithiophosphate nanosheets (Ni_1−xFe_xPS₃ samples, x = 0, 0.05, 0.1, 0.15, 1) could be synthesized by a traditional CVT method, followed by liquid exfoliation. The resulting FePS₃ was almost inactive for OER catalysis, whereas the Ni_1−xFe_xPS₃ nanosheets showed much better OER performance than the undoped NiPS₃. Ni_0.9Fe_0.1PS₃ exhibited the best OER activity among all the Ni_1−xFe_xPS₃ samples, providing a current density of up to 20 mA cm⁻² at an overpotential of 329 mV. In comparison, our group developed a novel solid-state transformation (SST) process for the mass production of MPQ₃ nanosheets [80]. Using this method, Fe-doped NiPS₃ nanosheets (Fe = 5.93% in atomic percentage) with a smaller lateral size (~ 100 nm) than the exfoliated congeners (several micrometers) could be prepared. Interestingly, the OER performance of the Fe-doped NiPS₃ nanosheets was much better than that of the NiPS₃ and Fe-doped Ni(OH)₂ nanosheets, as well as that of the exfoliated Ni_0.9Fe_0.1PS₃ nanosheets. The optimized sample showed an overpotential of 256 mV to reach a current density of 30 mA cm⁻². The Tafel slope of this Fe-doped NiPS₃ nanosheet (46 mV dec⁻¹) was also better than that of IrO₂ (56 mV dec⁻¹) and RuO₂ (86 mV dec⁻¹). As demonstrated by the ECSA tests, electrochemical impedance spectroscopy (EIS), and DFT calculation, the superior electrocatalytic activity of the Fe-doped NiPS₃ nanosheets might be ascribed to the Fe-doping. DFT calculation revealed that doping with Fe could improve the electronic conductivity and significantly weaken the interaction between NiPS₃ and the oxygen-containing intermediates (Fig. 6). Combined with all of the above-mentioned theoretical studies related to electrocatalytic water splitting, metal doping in MPQ₃ materials is a prospectively efficient way to achieve intermediate binding energy of hydrogen/oxygen species (and, therefore, improved catalytic performance). In this regard, the integration of metals that fall on the opposite slopes of the volcano plots should be reasonable for the design of multicomponent MPQ₃ catalysts [7, 81].

Open image in new window

Furthermore, we also designed a series of novel 0D–2D nanohybrids composed of a Ni_1−xFe_xPS₃ (x = 0.1, 0.2, 0.3, 0.4) nanomosaic decorated on the surface of Ti₃C₂T_x@MXene [82]. Here, nanomosaic refers to Ni_1−xFe_xPS₃ nanoparticles with irregular shapes that are dozens of nanometers in lateral size, which covered the surface of MXene to form a mosaiclike structure. Such a heterostructure integrating the Ni_1−xFe_xPS₃ nanomosaic and conductive Ti₃C₂T_x MXene can combine the advantages of each constituent to significantly enhance the conductivity and active sites. The resulting nanohybrids exhibited improved HER and OER performance relative to the individual materials. Among all the tested samples, the Ni_0.7Fe_0.3PS₃@MXene nanohybrid, as an OER catalyst, showed excellent catalytic performance (282 mV@10 mA cm⁻²) and a Tafel slope of 36.5 mV dec⁻¹. Further, the Ni_0.9Fe_0.1PS₃@MXene nanohybrid showed optimized HER performance in alkaline solution (with an overpotential of 196 mV). Due to their efficacy for the electrocatalytic HER and OER, overall water splitting could be achieved with the nanohybrid-based catalysts. The Ni_0.7Fe_0.3PS₃@MXene||Ni_0.9Fe_0.1PS₃@MXene couple showed a low onset potential of 1.42 V and a potential of only 1.65 V to reach a current density of 10 mA cm⁻².

In this minireview, recent advances in the study of metal trichalcogenidophosphates for photo- and electrocatalytic water splitting were briefly summarized. Theoretical calculations have revealed the possibility of adjusting the electronic structure of MPQ₃-based materials through composition tuning, doping, and heterostructure interface conjunction. The success of MPQ₃ catalysts with various compositions and structures for overall water splitting has confirmed their promising practical application. There is no doubt that MPQ₃-based materials are good candidates as high-performance photo- or electrocatalytic water splitting catalysts. However, studies on the catalytic properties of MPQ₃ are still in the early stage, and there are still many challenges related to MPQ₃ that must be overcome, such as preventing anodic corrosion during the photocatalytic water splitting process, proper selection of doping metals, probing the active site, and exploring the catalytic mechanism during the HER and OER processes.

In the near future, the most urgent task related to MPQ₃ catalysts is understanding the relationships among the composition, electronic structure, and adsorption free energies of the reactive intermediates on the MPQ₃ surface. Systematic understanding of doping metal selection is needed. Moreover, the construction of MPQ₃-based heterostructures is another important undertaking for achieving high-performance water splitting catalysts. In this regard, more effort should be devoted to control of the carrier separation and migration in MPQ₃-based heterostructures by carefully selecting the co-catalysts.