2D MXenes as Co-catalysts in Photocatalysis: Synthetic Methods

Highlights Two-dimensional transition metal carbides/nitrides (MXenes) as co-catalysts were summarized and classified according to the different synthesis methods used: mechanical mixing, self-assembly, in situ decoration, and oxidation. The working mechanism for MXenes application in photocatalysis was discussed. The improved photocatalytic performance was attributed to enhancement of charge separation and suppression of charge recombination.


Introduction
Energy shortage and environmental pollution have become the two major issues faced by humanity due to limited fossil fuel resources and increasing consumption. Developing sustainable and clean energy is the key to addressing these two problems [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15]. In being clean and inexhaustible, solar energy shows great potential to be one of the most promising future energy sources. Solar energy can be exploited in photovoltaic technologies [16], CO 2 photoreduction [17,18], N 2 photo-fixation [19], degradation of organic compounds [20][21][22][23][24][25][26], and photocatalytic water splitting [27]. In renewable hydrogen fuel-based photocatalytic water-splitting systems [28][29][30], photocatalysts play a critical role [31,32]. Photo-catalyzed solar energy conversion can be divided into three steps: (1) light absorption, (2) charge separation and transfer, and (3) surface reaction. Any improvement on each of these steps will contribute to enhancing the total conversion efficiency. Conventional photocatalysts such as TiO 2 , g-C 3 N 4 , and CdS demonstrate low photocatalytic efficiency due to rapid charge recombination in these materials. Using noble metals such as Pt, Ru, and Pd as co-catalysts will increase cost, although such materials can enhance charge separation ability and suppress recombination of charges. A co-catalyst that is both efficient and cheap is thus urgently needed to promote the development of photocatalysis.
MXenes, comprising transition metal carbides, nitrides, and carbonitrides, are a new family of two-dimensional (2D) materials that have attracted much attention in recent years [2]. The general formula of MXene is M n+1 X n (n = 1, 2, 3), where M represents a transition metal, such as Sc, Ti, Zr, Hf, V, Nb, Ta, and Mo, while X represents C and/or N. Owing to their unique structure and superior photoelectronic properties, layered structure MXenes show various potential applications in different areas, such as energy storage [3,[33][34][35][36][37][38], electromagnetic interference shielding [39,40], gas sensors [41], wireless communication [42], water treatment [43,44], solar cells [45][46][47], and catalysis [41,[48][49][50][51]. 2D MXenes are being increasingly studied in the past few years, as evidenced by the rapidly increasing number of scientific articles published per year (Fig. 1a). MXenes are usually synthesized by selectively etching the A layer from MAX phases, which constitute a family of tertiary ductile ceramics, where the A layer is made of an element such as Al, Ga [52], or Si [53]. After selective etching of the A layer, 2D MX layers with surface functional groups (-O, -OH, -F, or a mixture of several groups denoted as T x ) are left. The most widely used methods for selective etching are wet chemical HF etching and in situ HF etching (using a mixture of acids and fluoride salts), although other routes using tetramethylammonium hydroxide (TMAOH) [54,55], electrochemical [56,57], or etching with NaOH [58], and ZnCl 2 [49]) have also been explored. Generally, multilayered MXenes are produced by HF etching, whereas single or few-layered MXene flakes are obtained by in situ HF etching or through delamination of a multilayered MXene by intercalation of large organic molecules (Fig. 1b). The etching methods of Ti 3 C 2 T x MXene, which is the first discovered and the most studied MXene, have been reviewed elsewhere [59,60].
In view of the rapid development in the application of 2D MXenes, several reviews on their synthesis [59][60][61], and application in energy storage [33,48,62] and catalysis  [51] have been reported. MXenes are promising for application in photocatalysis [63] because of their large surface area, good conductivity, presence of a sufficient number of active sites, and containing suitable elements for effective photocatalysis, but they cannot be directly used as photocatalysts since MXenes are generally not semiconductors [51,62]. Although there are some MXene semiconductors that have been predicted theoretically [64][65][66][67][68], these have not yet been experimentally synthesized. In this review, we give a detailed discussion on MXene as a co-catalyst in photocatalysis and describe the different methods used for the synthesis of MXene-derived photocatalysts, along with problems encountered in this system and a prospective outlook on future research in this field.

Synthetic Methods for MXenes as Co-catalysts in Photocatalysis
In view of their good conductivity and large surface area, MXenes have been applied in photocatalysis both to replace noble metal co-catalysts and to enhance the charge separation ability of the photocatalyst (Fig. 2). The most common methods used for the preparation of photocatalyst composites include mechanical mixing, self-assembly, in situ decoration and oxidation, or a combination of the three methods.

Mechanical Mixing and Self-assembly
Mechanical mixing is the easiest method to form photocatalyst composites. Stirring the two components in the liquid phase or grinding of powders can be used for sample preparation. Interestingly, due to electrostatic attraction, photocatalysts with positive charge are easily combined with MXenes whose surfaces are enriched with negative charges, leading to self-assembled photocatalyst composites. In addition, the self-assembling property could be further improved by using other induced techniques simultaneously, where the photocatalysts and co-catalysts are prepared in advance [44]. An et al. [72] demonstrated that synergetic effects of Ti 3 C 2 MXene and Pt when used as dual co-catalysts enhanced the photoactivity of g-C 3 N 4 for hydrogen evolution (Fig. 3a), where HF-etched exfoliated Ti 3 C 2 and g-C 3 N 4 were mixed in liquid by stirring followed by photodeposition of Pt on the composites. co-catalysts-modified photocatalysts (g-C 3 N 4 /Ti 3 C 2 /Pt) was much better than that of Pt-or Ti 3 C 2 -only systems, reaching 5.1 mmol h −1 g −1 in hydrogen production (Fig. 4a). This enhanced performance was due to the presence of Ti 3 C 2 MXene that facilitated interfacial charge separation and carrier transport from the conduction band (CB) of g-C 3 N 4 to Pt. Our group prepared g-C 3 N 4 /Ti 3 C 2 T x composites by grinding g-C 3 N 4 and Ti 3 C 2 T x powders together followed by annealing in different gas atmospheres, to tune the surface termination groups (Fig. 4b) [74]. X-ray photoelectron spectroscopy data showed an increase in -O termination groups accompanied by a decrease in -F termination groups on the surface of Ti 3 C 2 . Ti 3 C 2 with -O termination groups had better photoactivity, revealing that the presence of such groups in Ti 3 C 2 had a positive effect on hydrogen production by increasing the number of active sites. Moreover, this finding was consistent with density functional theory (DFT) simulation results. The |ΔG H | of Ti 3 C 2 with -O terminations was found to be as low as 0.01 eV, which is lower than that of Pt (111). In a similar study, Ye et al. [69] treated HF-etched Ti 3 C 2 with KOH to convert -F groups into -OH groups, and then combined the KOH-treated Ti 3 C 2 with TiO 2 (P25) powder by stirring in water (Fig. 3c). DFT calculations demonstrated that -OH groups played the role of active sites for the adsorption and activation of CO 2 reduction [69]. Experimentally, the photoactivities for CO 2 reduction were increased 3 times and 277 times after KOH treatment, for CO and CH 4 , respectively (Fig. 4d). Interestingly, increasing the number of -OH groups not only improved the photo-conversion efficiency but also changed the nature of the products. The -OH groups resulting from KOH treatment provided more active sites for CO 2 adsorption and enabled greater electron transfer to CO 2 and facilitated its reduction to CH 4 . Though the surface termination groups can be changed through annealing and KOH treatments, -F groups could not be completely exchanged. More studies to precisely tailor the termination groups need to be carried out in the future. Xie et al. [73] used an electrostatic self-assembly process to combine positively charged CdS nanosheets and Ti 3 C 2 nanosheets (possessing negative charge) (Fig. 3b) for CO 2 reduction (Fig. 4c). Cai et al. [75] synthesized Ag 3 PO 4 /Ti 3 C 2 by electrostatically driven self-assembly method, which had the advantage of being a mild method that prevented Ti 3 C 2 from oxidation. The composites showed better performance than reduced graphene oxide (rGO), and this preparation procedure provided a new direction to the preparation of semiconductor-MXene composites. Liu et al. [44] fabricated a 2D layered and stacked g-C 3 N 4 /Ti 3 C 2 composite by evaporation-induced self-assembly and used it to degrade organic pollutants (ciprofloxacin) (Fig. 3d). Both photogenerated holes and superoxide radicals (·O 2 − ) resulting from photogenerated electrons played important roles in ciprofloxacin decomposition (Fig. 4f); in this process, self-assembly was an efficient method that allowed intimate mixing of the components in the composite. The sample was also more homogeneous than mechanically mixed ones because of the electrostatic attraction between the charged entities. However, opposite charges on each surface were required for self-assembly, which limited wider application of this process. Therefore, other techniques to induce self-assembly such as evaporation-induced self-assembly were developed to widen the range of application of products [44]. The above-mentioned MXene-based composites prepared by mechanical mixing and self-assembly methods for photocatalysis application are summarized in Table 1. Results from all these works prove that 2D MXene is an efficient additive material to enhance charge separation and charge transfer during photocatalysis. In these two methods, the properties of MXenes are retained by avoiding high temperature and use of other solvents or surfactant. No change in oxidation or surface termination groups occurs in these synthesis methods. Therefore, these two are the easiest and allow synthesis under the mildest conditions.

In Situ Decoration of Semiconductors onto the Surface of MXenes
In contrast to composites prepared by mechanical mixing of materials, in situ decoration methods consist in synthesizing a different material directly onto the MXene surface. As a result, in situ synthetized materials and MXenes are chemically bonded, which could be an important advantage in some designs. However, the range of viable synthetic conditions for in situ decoration is limited, because MXenes are easily oxidized in solution, especially at high temperatures [107]. It is therefore necessary to use mild conditions to protect MXenes from oxidation, especially when monoand few-layered MXenes are used. So far, g-C 3 N 4 , TiO 2 , CdS, and bismuth compounds have been bonded to various MXenes using this strategy. g-C 3 N 4 is one 2D semiconductor material that is combined with MXenes used as a co-catalyst in the photocatalysis process (Fig. 5). MXene can be added during the calcination of a precursor, such as melamine and thiourea, but the high calcination temperature (around 550 °C) may cause the oxidation of MXene into TiO 2 . The high photoactivity of g-C 3 N 4 /MXene is attributed to the efficient charge separation; moreover, the heterojunction formed by TiO 2 /g-C 3 N 4 also plays an important role in charge separation [108]. Shao et al. [81] synthesized Ti 2 C/g-C 3 N 4 by melamine calcination and used it in hydrogen production (Fig. 5a, d). Though the ratio of Ti 2 C in the composite was as low as 0.4 wt%, a peak due to TiO 2 resulting from the oxidation of Ti 2 C could be seen in the XRD pattern. Liu et al. [19] synthesized TiO 2 @C/g-C 3 N 4 heterojunction by melamine calcination (Fig. 5b), where Ti 3 C 2 was oxidized to TiO 2 @C during the calcination process. This composite was highly effective in the reaction of nitrogen reduction to ammonia, with the best performance reaching as high as 250.6 μmol h −1 g −1 , which was better than that of TiO 2 @C and g-C 3 N 4 (Fig. 5e). Xu et al. [82] synthesized Ti 3+ -rich Ti 3 C 2 /g-C 3 N 4 by calcination of thiourea and employed it as an electrode for CO 2 reduction in a photoelectrocatalytic (PEC) system (Fig. 5c, f), achieving a total CO 2 reduction rate of 25.1 mmol h −1 g −1 .
The Ti 3+ species suppressed charge recombination at the Ti 3 C 2 /g-C 3 N 4 heterojunctions, leading to a corresponding increase in CO 2 conversion efficiency.
Apart from the above-mentioned synthesis methods, composite photocatalysts can also be synthesized by combining TiO 2 , a metal sulfide, or a bismuthide with MXene under hydrothermal conditions (Fig. 6). Gao et al. [83] synthesized TiO 2 /Ti 3 C 2 nanocomposites by a hydrothermal method using TiSO 4 as a precursor for methyl orange (MO) degradation (Fig. 6a), where small TiO 2 particles could be observed on the surface of multilayered Ti 3 C 2 . Wang et al. [84] employed TiCl 4 as the precursor in the hydrothermal synthesis of rutile TiO 2 /Ti 3 C 2 T x for hydrogen production by water splitting (Fig. 6d). The photocatalytic activity of TiO 2 when combined with other MXenes (Ti 2 CT x and Nb 2 CT x flakes) as co-catalysts was also explored; results proved that in general, MXenes could be used as effective co-catalysts for solar hydrogen production. Ran et al. [70] combined CdS and Ti 3 C 2 particles by a one-step hydrothermal reaction (Fig. 6b). A hydrogen production rate of 14,342 μmol h −1 g −1 was achieved when using Ti 3 C 2 as the co-catalyst; this performance is 136.6 times higher than that of the pure CdS photocatalyst. The effectivity and versatility of Ti 3 C 2 MXene as a co-catalyst for photocatalytic hydrogen production was demonstrated by other metal sulfides (ZnS) [91] photocatalysts as well. Xie et al. [73] showed that Ti 3 C 2 flakes enabled the local confinement of Cd 2+ released during photo-corrosion and thus enhanced the stability of the metal sulfide. Besides CdS, In 2 S 3 /Ti 3 C 2 T x hybrids synthesized by hydrothermal method have been used for methyl orange degradation as reported by Wang et al. [90]. Among the hybrids based on other additives (carbon nanotubes (CNT), rGO, MoS 2 , and TiO 2 ), Ti 3 C 2 -based composites showed the best photocatalytic activity, which is attributed to their high electrical conductivity. Shi et al. [85] synthesized TiO 2 /C/ BiVO 4 composites by hydrothermal method for the degradation of Rhodamine B, where Ti 3 C 2 was employed both as a support for the growth of BiVO 4 nanoparticles and as a precursor for the generation of 2D-carbon upon oxidation. The electron transfer process was accelerated by the presence of Ti 3 C 2 -derived 2D-carbon layers, thus improving the photocatalytic performance for Rhodamine B degradation. Ultrathin 2D/2D heterojunction of MXene/Bi 2 WO 6 prepared by the in situ growth of ultrathin Bi 2 WO 6 nanosheets on the surface of ultrathin Ti 3 C 2 nanosheets for photocatalytic CO 2 reduction was reported by Cao et al. [88] (Fig. 6c). The CH 4 and CH 3 OH yield were 4.6 times higher than those obtained with pristine Bi 2 WO 6 , which was ascribed to the enhanced CO 2 adsorption arising from the increased specific surface area and improved pore structure of the layered heterojunction. The different composites/hybrids containing MXene or MXene-derived products prepared by hydrothermal methods and used in photocatalysis are listed in Table 1. The synthetic process for MXenes-based composites includes doping into the photocatalysts or using MXene as a support for in situ decoration of the semiconductor photocatalyst. The chemical reactions taking place during photocatalyst formation led to increased interfacial area, thus providing greater possibilities for the transfer of photogenerated electrons. However, one disadvantage of this method is the oxidation of MXenes during photocatalyst synthesis. Although difficult to precisely characterize, conditions of formation of the photocatalysts may be too harsh and cause structural degradation of MXenes, especially in the case of single-layered MXenes, due to their lower stability toward oxidation.

MXene-Derived Photocatalysts
Different from mechanical mixing, self-assembly, and decoration methods, the in situ oxidation method using MXene (Ti 3 C 2 is the most studied example) as a precursor for the synthesis of photocatalysts has also been explored (Fig. 7). Peng's group tuned the facet of TiO 2 /Ti 3 C 2 using a hydrothermal method without using an additional TiO 2 precursor (Fig. 7a, b) [71,93]. NaBF 4 and NH 4 F were used as reagents to, respectively, control morphology in the synthesis of (001) TiO 2 /Ti 3 C 2 and (111) TiO 2 /Ti 3 C 2 , which were then applied in methyl orange degradation. Both the facet type of TiO 2 and the ratio of TiO 2 to Ti 3 C 2 could be controlled by changing the duration of the hydrothermal reaction. Jia et al. [94] obtained closely aggregated TiO 2 nanorods with high carbon doping starting from Ti 3 C 2 flakes and demonstrated a better photoactivity than commercially available P25 for hydrogen production (Fig. 7c). The carbon doping also changed the electron structure of TiO 2 and enhanced its light absorption ability. Peng et al. [95] also used Ti 3 C 2 as a hole trap and Cu as an electron trap to separate the charges through a dual-carrier-separation mechanism, showing the potential of MXene as an efficient functional material for photocatalysis (Fig. 7d). Calcination under atmosphere containing gases such as CO 2 and O 2 is another method used for the controlled oxidation of MXenes (Fig. 8). Lu et al. [96] obtained Ti 3 C 2 / TiO 2 /CuO by annealing Cu(NO 3 ) 2 and Ti 3 C 2 together under argon atmosphere (Fig. 8a). Because of its good electronic conductivity, the incorporation of Ti 3 C 2 improved electron/ hole separation and led to better methyl orange degradation. Yuan et al. [97] annealed Ti 3 C 2 in CO 2 to prepare 2D-layered C/TiO 2 hybrids used in hydrogen production, in which the presence of 2D carbon layers increased electron transport channels and enhanced charge separation efficiency (Fig. 8b). In addition, the effects of oxidation temperature and CO 2 on the grain size and crystal structure of TiO 2 were also investigated, revealing that increasing oxidation temperature and CO 2 gas flux led to larger grain sizes and more rutile TiO 2 formation. Low et al. [98] calcined Ti 3 C 2 at different temperatures, enabling the in situ growth of TiO 2 nanoparticles on Ti 3 C 2 nanosheets, thus forming TiO 2 /Ti 3 C 2 composites with different loading amounts of TiO 2 with the aim to improve performance in CO 2 reduction reaction (Fig. 8c). Interestingly, three main products were obtained during the photocatalytic CO 2 reduction process due to the sufficiently high intrinsic reduction potential of TiO 2 .
Results of the study also pointed out that excess of Ti 3 C 2 in the composite could have an adverse effect on photocatalytic performance. Su et al. [99] used CO 2 to partially oxidize Nb 2 C to form Nb 2 O 5 /Nb 2 C composites for hydrogen production, where Nb 2 O 5 and metallic Nb 2 C served, respectively, as the semiconductor photocatalyst and co-catalyst (Fig. 8d).
The easily formed junction at the interface served as an electron sink to efficiently capture photogenerated electrons and suppress recombination of photogenerated electron-hole pairs, thus enhancing the efficiency of charge separation and contributing to improved photocatalytic activity [71,93,99,102]. Besides the hydrothermal method and calcination, other routes such as chemical oxidization and high-energy ball milling were also used to oxidize MXenes (Fig. 9). Cheng et al. [100] oxidized Ti 3 C 2 flakes with 30% H 2 O 2 to form microporous-MXene/TiO 2−x nanodots (Fig. 9a). This composite worked as a photo-Fenton bifunctional catalyst for Rhodamine B degradation under both dark and illumination conditions. Li et al. [101] synthesized TiO 2 @C nanosheets from Ti 2 C by high-energy ball milling and used it for methylene blue degradation (Fig. 9b). Shortly thereafter, our group used water to oxidize Ti 3 C 2 to be applied in hydrogen  [95]. Copyright 2018 Elsevier production using Eosin Y as a sensitizer [102]. Similar to other oxidized MXenes, amorphous carbon and TiO 2 were formed after oxidation (Fig. 9c, d). The various MXenederived composites obtained by in situ oxidation to be used as photocatalysts are listed in Table 1.
The MXenes oxidation is different from other methods because of the residual presence of carbon (mostly amorphous carbon) after oxidation, and the M element is oxidized into metal oxide on the carbon layer. Thus, the composite obtained is of the form metal oxide/MXenes/C. Both MXenes and C can be used as co-catalysts in the photocatalysis process. However, in this method, the ratio of the photocatalyst to MXenes varies within a certain range since no precursor is introduced. The limitation of this method is that only a few semiconductors (depending on M element) can be used as the photocatalyst.

Mechanism of MXenes as Co-catalysts
Since MXenes are conductors and serve as co-catalysts, the mechanism of action of a MXenes-based photocatalytic system is through accelerated charge separation and suppression of carrier recombination [69][70][71]. The photocatalysts absorb visible light and photogenerated electrons are excited to the CB, while holes are left in the valence band (VB). The excited charge carriers are transferred to MXenes at the interface mainly because of the higher potential of MXenes. Electrons transfer to MXenes without recombination and react on the MXene surface to generate H 2 by reducing H + [74,78,81,91,94,102,103], CH 4 and CO by reducing CO 2 [88,98], or NH 3 by reducing N 2 [19], as shown in Fig. 10 process (a). In process (b), holes transfer to MXenes and react to produce OH· that can be utilized for degradation of organics [71,93,95]; electrons can also produce OH· for organic degradation [71,93]. The charge transfer process  from the photocatalyst to MXenes improves electron-hole pair separation and suppresses charge recombination in photocatalysts, thus enhancing the photoactivity.
Another advantage of using MXenes in photocatalysis is due to their termination groups. For example, -O termination groups show the best potential for hydrogen production because of their low |ΔG H | and the availability of active sites for the adsorption of hydrogen atoms [70,74].
Though termination groups are important in photocatalysis, currently, it has not been possible to precisely control the relative concentrations of the different termination groups. Using presently available synthetic methods, changing the different reaction conditions can partially modify the termination groups on MXenes surface and thereby affect their performance in photocatalysis.

Conclusion and Outlook
In summary, the application of MXenes in photocatalysis has shown rapid development since 2015. Among the MXenes family, Ti 3 C 2 has been the most studied MXene. Mechanical mixing and self-assembly are mild and easy methods of synthesis, where the ratio of MXenes to the photocatalyst can be controlled. In addition, MXenes can also be doped into the photocatalysts by in situ decoration of a semiconductor photocatalyst. The large interfacial area afforded by   the doping process improves electron transfer. However, the MXenes oxidation method has the advantage of obtaining both carbon and MXenes as co-catalysts by forming a metal oxide/MXenes/C structure. Though the above-mentioned four synthetic methods are generally used for photocatalysts, with further development in the field of MXenes, new processes may be discovered. Besides developing improved synthetic methods, the other aspects that need to be focused on in the future are as follows:  [34,109] and have shown potential for photocatalysis applications. More work needs to be done in this direction.

5.
Developing new synthesis methods for MXenes. HF and in situ HF wet chemical treatment are by far the most used methods in MXenes synthesis. Other HF-free methods are emerging and leading to MXenes with different properties. Yet, these have not been investigated in photocatalytic applications, and thus, the effect of the type of synthesis process used on the final performance of the MXene is currently not understood.
In short, due to tremendous effort of scientists worldwide, the great potential of MXenes in photocatalysis has been revealed. With the fast-growing development in this area, it is expected that more and more studies will focus on the applications of MXenes photocatalysis and pave the way to the commercialization of photocatalytic technologies based on these materials.