Two-dimensional carbide/nitride (MXene) materials in thermal catalysis

Abstract

Two-dimensional carbide/nitride materials, i.e., MXenes, were first synthesized from the corresponding MAX phases in 2011. Since their discovery, they have been widely applied in batteries, supercapacitors, electromagnetic shielding materials, electrocatalysis, and photocatalysis due to their unique and tunable physical, chemical, and electrical properties. Recently, MXenes have been applied in thermal catalytic reactions, such as hydrogenation, dehydrogenation, water–gas shift reactions, and desulfurization due to their thermal stability and superior catalyst properties similar to noble metals. In this article, we systematically summarize the characteristics of MXenes as catalysts and supports compared with other traditional thermal catalysts with respect to both structures and catalytic activities. Furthermore, the nature of termination groups, active sites, and metal–MXene interactions are elaborated. Finally, we provide insights into the future development of catalysts based on MXene materials.

---

Hui Zhou

Hui Zhou is an assistant professor at the Department of Energy and Power Engineering, Tsinghua University, China. His research interests include 2D materials, heterogeneous catalysis, bioenergy, and CO2 utilization. He has published more than 60 peer-reviewed journal articles with more than 2500 citations. He was a Marie Curie Individual Fellow at ETH Zurich. He has also been awarded the Top Ten Rising Stars of Science and Technology of China. He is currently the Founding Associate Editor of Carbon Capture Science & Technology and the Associate Editor of Frontiers in Energy Research.

---

1. Introduction

Two-dimensional (2D) materials have unusual chemical and physical properties in contrast with traditional bulk materials due to their unique planar structures.^1–4 Therefore, increasing types of 2D materials with different structures and compositions were discovered, such as, graphene,^5,6 transition metal dichalcogenides (TMDs),^4,7 hexagonal boron nitride (h-BN),³ layered double hydroxides (LDHs),⁸ and 2D metal–organic frameworks (MOFs).⁹ Among them, MXenes, first synthesized in 2011,^10–12 have gradually become the hot spot in the library of 2D materials. MXenes are two-dimensional transition metal carbides, carbonitrides, and nitrides with the general formula of M_n+1X_nT_x (n = 1–3), where M is an early transition metal (such as Sc, Ti, Zr, Hf, V, Nb, Ta, Cr, and Mo), X is carbon and/or nitrogen, and T_x stands for surface terminations (hydroxyl, oxygen, or fluorine)_.^13–15 Till now, more than thirty kinds of MXenes with different stoichiometric compositions have been synthesized; countless possible solid solutions of MXenes have been designed theoretically; besides, plenty of new types are still under investigation.^16–18 Due to their diverse elemental compositions and adjustable surface terminations,^17,19 MXenes possess various interesting chemical and physical properties like a tunable bandgap,^1,20 good electrical conductivity,^21–24 excellent photoelectronic properties, and ultra-low optical attenuation.²⁵ Based on these superior features, MXenes play a significant role in many fields, such as energy storage and conversion,^26–28 sensors,^18,29 electromagnetic interference shielding,³⁰ absorption,³¹ electrocatalysis,^32–34 and photocatalysis.^23,35,36 Recently, MXenes have also been sought-after in thermal catalysis, due to their tunable surface terminations and high thermal stability.^37,38 Moreover, it has been reported that MXenes have similar catalytic properties to noble metals.^39,40

In thermal catalysis, MXenes can work as both catalysts and supports (Fig. 1). When MXenes serve as catalysts, they play a decisive role in many typical thermal reactions including hydrogenation,¹⁶ dehydrogenation,⁴¹ CO oxidation,^18,38 water–gas shift reaction,⁴² and N₂ fixation,⁴³ in which Ti₃(C/N)₂T_x, Mo₂CT_x, and V₂CT_x are applied extensively.^44–46 When MXenes are used as supports, Ti₃(C/N)₂T_x and Mo₂CT_x are widely used, and transition metals are loaded on them in the forms of nanoparticles (NPs),⁴⁷ nanoclusters (NCs),⁴⁸ or single atoms (SAs).⁴⁹ Moreover, to further enhance the catalytic performance,⁵⁰ some efforts, like surface functionalization,⁵¹ heteroatom doping, and defect creation,³⁷ have also been made.⁵²

Fig. 1 The application of MXenes in thermal catalysis.

This review summarizes the application of MXene materials in thermal heterogeneous catalysis. We discuss MXenes as both catalysts and supports, with an emphasis on the termination groups, active sites, and the interactions of MXenes with transition metals (Fig. 1). Finally, we discuss the future opportunities and challenges for MXene-based catalysts in this field.

2. Structure and synthesis of MXenes

2.1 Structure of MXenes

The structure of MXenes includes the M_n+1X_n skeleton from the corresponding MAX phases (precursors of MXenes) and the interfacial termination groups from the specific synthetic process.³⁴ The MAX phases are layered ternary carbides and nitrides,¹³ with the hexagonal structure of M_n+1X_n units and alternately stacked A atoms (mainly elements of groups 13 and 14 of the periodic table, such as Al and Si).²⁵ In ceramic phases of MAX, the “M–X” interactions are mainly covalent and ionic bonds, whereas “M–A” bonds are metallic.¹³ As shown in Fig. 2, MXenes inherit the hexagonal atomic lattice P6₃/mmc from the MAX parent phase, in which the M atoms are hexagonally closed packed and X atoms fill the octahedral interstitial sites.¹ Through the change of n (Fig. 2(i)) in M_n+1X_n units, M_1.33X, M₂X, M₃X₂, and M₄X₃ can be obtained. Moreover, by varying the type of M element in structural motifs (Fig. 2(ii)), solid-solution M_n+1X_n units and o-M_n+1X_n (a type of ordered MXene) are fabricated.¹⁷

Fig. 2 The structure of MXenes with different M_n+1X_n units and termination groups, where M refers to an early transition metal, X is carbon and/or nitrogen, and n stands for the number of X.

Among mono-M MXenes, M atoms in the M₂XT_x phase follow ABABAB ordering corresponding to hexagonal close-packed stacking, whereas M atoms are face-centered cubic stacked with ABCABC ordering in M₃C₂T_x and M₄C₃T_x phases.⁵³ Recently, i-MXenes with the formula of M_1.33XT_x were discovered. The materials have ordered vacancies and only one kind of M element dispersed in the basal plane.⁵⁴ Furthermore, MXenes contain two kinds of M elements in their components, which greatly expand their variety. More than ten kinds of double-M MXene phases have been synthesized and thousands have been predicted (Fig. 3). In the solid solution ((M,M′)_n+1C_nT_x), M and M′ metal elements are randomly distributed throughout the entire structure,¹⁷ and their electrochemical⁵⁵ and catalytic properties can be tuned by controlling the ratio of these two transition metal elements.⁵⁶ Up to now, many solid solutions, such as (Ti,V)₂CT_x, (Ti,Nb)₂CT_x, (Mo,V)₄C₃T_x, and (Nb,Zr)₄C₃T_x, have been synthesized; however, they account for only a small part of countless predicted phases.¹⁷ As for ordered MXenes, o-MXenes are one of the most important types, where one or two layers of an M element are sandwiched between layers of another M element.⁵⁷ It has been reported that in synthesized o-MXenes, the outer layers are commonly occupied by Mo or Cr, while the center is occupied by Ti or V.¹

Fig. 3 The predicted (in black) and synthesized (in blue) MXenes classified by different types.

In addition to the M_n+1X_n skeletons, the electronegative termination groups, formed in the aqueous synthetic system, also play an important role in providing MXenes with excellent properties. To stabilize the MXenes, termination groups are usually located at three sites including (i) the top of the transition metal atoms, (ii) the hollow position between the top metal atoms, and (iii) the hollow position between the next stack of X atoms. Due to the low steric hindrance surface, terminations located at site (i) on both sides of MXene sheets are the most stable configuration for most MXenes according to density functional theory (DFT) calculations. However, site (ii) becomes more energetically favorable for MXenes when the transition metals cannot provide sufficient electrons to both X and surface terminations. For example, O-termination MXenes with low valency transition metals tend to be either in site (ii) configuration or mixed sites (i) and (ii), because two electrons are required to stabilize their adsorption position, but only one electron can be obtained from transition metal surface sites.¹¹ Furthermore, changing the coverage of surface termination groups can have an influence on the selectivity and activity of the reactions.⁵⁸

2.2 Synthesis of MXenes

MXenes are commonly synthesized by a top-down etching process in which “A” layers are selectively etched from their corresponding MAX phases.¹⁷ The “M–A” metallic bonds are more active in contrast with covalent/ionic/metallic bonds between M and X.^13,59,60 Hydrofluoric acid (HF) was the first used etchant to synthesize MXenes (Fig. 4a).⁶¹ In 2011, Gogotsi et al. found that the Al atomic layers in Ti₃AlC₂ can be selectively etched using 50% HF due to the high reactivity between the Al-containing MAX phase and F⁻, resulting in accordion-like Ti₃C₂T_x powder.⁶¹ After the synthesis of Ti₃C₂T_x by this method, many Al-containing MAX phases, e.g., V₄AlC₃, V₂AlC, Mo₂TiAlC₂, Ta₄AlC₃, and Nb₂AlC, were etched with HF to produce the corresponding MXenes. However, HF is a type of toxic and hazardous reagent, which threatens the environment and the human body. Researchers are exploring milder and safer ways to synthesize MXenes.^1,62 One method is to produce HF in situ, and the typical reaction is that HF can be formed in the reaction between hydrochloric acid and metal fluorides such as LiF and KF.^63,64 Besides, HF can also be synthesized through the hydrothermal process of ammonium fluoride and hydrolysis of ammonium bifluoride solution (Fig. 4b).^65,66 After reacting with fluorine-containing compounds, the obtained MXenes are always decorated with many surface terminations such as –OH, –O, and/or –F.⁶⁷ By using this method, many carbide MXenes have been synthesized, such as Ti₃C₂T_x,⁶³ Nb₂CT_x,^64,68 and Zr₃C₂T_x.⁶⁹

Fig. 4 Schematic diagram of the top-down selective etching method for preparing MXenes. (a) The exfoliation process of replacing an Al atom with –OH in the reaction of Ti₃AlC₂ and HF. Reproduced with permission.⁶¹ Copyright 2011, Wiley-VCH. (b) The selective etching of OH-terminated Ti₃C₂ in ammonium bifluoride solution. Reproduced with permission.⁶⁶ Copyright 2014, American Chemical Society. (c) The synthesis of Ti₄N₃T_x by molten fluoride salt treatment of Ti₄AlN₃ at 550 °C under Ar, followed by delamination of the multilayered MXene by tetraethyl ammonium hydroxide (TBAOH). Reproduced with permission.⁷⁰ Copyright 2016, Royal Society of Chemistry. (d) The element replacement reaction between the Lewis acid molten salt (ZnCl₂) and MAX phase to produce Cl-terminated MXenes. Reproduced with permission.⁷¹ Copyright 2019, American Chemical Society. (e) The synthesis of Cl-terminated MXenes by etching Ti₃AlC₂ precursors using KOH solution. Reproduced with permission.⁷² Copyright 2017, American Chemical Society. (f) Etching method based on the Ti₃AlC₂ anode and binary aqueous electrolyte. Reproduced with permission.⁷³ Copyright 2018, Wiley-VCH. (g) UV-induced selective etching route of removing the double Ga layers from Mo₂Ga₂C precursors to generate Mo₂C. Reproduced with permission.⁷⁴ Copyright 2020, Elsevier.

Additionally, using molten fluorides as etchants with the assistance of high temperature is an effective approach to acquiring MXenes, especially for those with high formation energy such as nitride MXenes.^70,75 For example, Gogotsi et al. first reported the synthetic process of 2D transition metal nitride (Ti₂N₃), in which a molten fluoride salt was used to etch Al from the Ti₄AlN₃ precursors at 550 °C under an argon atmosphere (Fig. 4c).⁷⁰ More recently, researchers paid more attention to fluorine-free ways to produce MXenes, including Lewis acid molten salt etching,^71,76,77 alkali etching,^72,78,79 and electrochemical etching.^80,81 Among them, Lewis acid molten salt etching is one of the most promising methods. Huang et al. designed a route in which the Zn element in molten ZnCl₂ replaced the Al element in MAX phase precursors, resulting in the synthesis of Ti₃ZnC₂, Ti₂ZnC, Ti₂ZnN, and V₂ZnC.⁷¹ Surprisingly, when ZnCl₂ was excessively used in this fabrication route, Cl-terminated MXenes were obtained (Fig. 4d). Therefore, the Lewis acid molten salt etching route was also used to modify the surface termination and etch non-Al MAX phases.^76,77 An alkali is also a kind of feasible etchant owning to the strong binding ability between hydroxyl from the etching solution and the amphoteric element Al of MAX.^67,78 Xie et al. put forward a two-step etching process in which Ti₃AlC₂ was first soaked in NaOH solution, and then hydrothermally treated using H₂SO₄ to obtain Ti₃C₂T_x.⁸² In this process, NaOH was used to leach the Al layer selectively, and H₂SO₄ helped to remove surface-exposed Al. Li et al. successfully fabricated single-layered Ti₃C₂(OH)₂ nanosheets by etching Ti₃AlC₂ precursors with KOH in the presence of a small amount of water (as shown in Fig. 4e).⁷² However, alkali etching synthesis methods are still facing some difficulties like complex separation operations, high alkali concentration, and strict temperature requirements, thereby leading to a low MXene yield.⁷⁹

As for the electrochemical etching method, Al atomic layers can be selectively removed under a certain voltage using MAX phases as the electrode with NaCl, HCl, or HF as the electrolytic solution.^67,83 The advantage of this approach lies in that the structure of MXenes can be precisely controlled by accurately managing the etching voltage window in the range of reaction potentials between A and M layers as well as regulating the appropriate etching time.^67,84 However, it is unavoidable that the parent MAX phases are often over etched and generate concomitant products, i.e. carbide-derived carbon (CDC),⁸¹ which decreases the yield of MXenes. Therefore, many researchers made efforts to overcome this drawback.⁸⁰ For instance, Yang et al. demonstrated an efficient etching method based on a Ti₃AlC₂ anode and binary aqueous electrolyte (Fig. 4f).⁷³ To ensure a continuous etching process and avoid the production of CDC, some intercalators such as TMAOH were used to enhance the accessibility of electrolyte ions into the Al layer. In addition to the etching methods mentioned above, recently, other methods like using a halogen as an etchant⁸⁵ or applying mechanical and electromagnetic waves to assist etching have also been reported. For example, Mei et al. proposed an effective UV-induced selective etching method without fluorine involved to fabricate mesoporous Mo₂C MXene from Mo₂Ga₂C (Fig. 4g).^67,74

After etching, stacked multi-layer MXenes, connected by van der Waals forces and hydrogen bonding, are usually obtained, which have fewer advantages compared to mono-layered MXenes.¹² Therefore, delamination is necessary to get single layers with high surface space, good hydrophilicity, and rich surficial functional groups.^86,87 A mechanical process like ultrasonic treatment was firstly used (Fig. 4a and g), while generally low yields were achieved due to the inability to disrupt strong interlaminar interactions. When metal halides such as LiF and KF were used in the preparation of MXenes, it has been found that metal ions can enter the space between the layers, leading to the increase of the layer space.⁶³ In this case, metal halides are excellent inorganic intercalators, allowing etching and exfoliation to occur simultaneously. Up to now, new organic intercalators, including dimethyl sulfoxide (DMSO),⁸⁸ TBAOH (Fig. 4c),⁸⁹ and tetramethylammonium hydroxide (TMAOH) (Fig. 4f),⁹⁰ have been developed to meet the demand for more efficient delaminating.⁸⁶

3. MXenes as catalysts

MXenes are an emerging group of materials in the field of thermal catalysis (Table 1). Due to the high thermal stability^91,92 and unique structures,⁴⁴ MXenes show some advantages compared with traditional catalysts. Among the numerous kinds of MXenes, Ti₃(C/N)₂T_x, Mo₂CT_x, and V₂CT_x have been intensively studied because of the relatively mature preparation process and the easy modification of their structures. Therefore, in this section, we describe the thermal catalytic reactions that occur on the above-mentioned MXenes (Ti₃(C/N)₂T_x, Mo₂CT_x, and V₂CT_x) as shown in Fig. 5, and compare MXenes with other catalysts from structures to activities. Furthermore, we summarize the specific roles of surface termination groups and propose the active sites on MXenes in the thermal catalytic reactions.

Table 1 The catalytic performance of MXenes

Catalyst	Reaction	Main reactant	T (°C)	P (MPa)	Conv.^a (%)	Target product	Selec.^b (%)	Ref.
a Conversion refers to the conversion of the main reactant. b Selectivity refers to the selectivity of the target product. c DBT is dibenzothiophene. d DBTO2 is dibenzothiophene sulfone.
Ti3C2Tx	Furfural hydrogenation	Furfural	180	5.0	36	Furfuryl alcohol	52	93
Ti3CNTx	Furfural hydrogenation	Furfural	180	5.0	46	Furfuryl alcohol	49	93
Ti3C2Tx	Guaiacol hydrodeoxygenation	Guaiacol	350	5.0	56	Phenol	61	94
Ti3C2Tx	Ethylbenzene dehydrogenation	Ethylbenzene	550	0.1	21	Styrene	97.5	95
Ti3C2Tx	HCOOH dehydrogenation	HCOOH	80	0.1	94	H2	100	41
Ti3C2	DBT^c oxidative desulfurization	DBT^c	130	0.1	95.7	DBTO2^d	—	96
Ti3C2/Ti3AlC2	DBT^c oxidative desulfurization	DBT^c	130	0.1	99.0	DBTO2^d	—	44
2D-Mo2C	CO2 hydrogenation	CO2	430	2.5	38	CO	94	51
V2CTx	CH4 dehydroaromatization	CH4	700	0.1	11.8	C6H6	4.84	97
2D-Mo2COx/SiO2	CH4 dry reforming	CH4	800	1.0	80	CO, H2	CO : H2 ∼ 1.5	98
Multilayered-V2CT	CH4 dry reforming	CH4	800	0.1	78	CO, H2	CO : H2 ∼ 0.9	99
Mo2CTx	Water–gas shift reaction	CO	500	0.1	18	CO2, H2	>99	100

---

Fig. 5 Commonly used MXene species as catalysts and their applications in thermal catalysis.

3.1 Ti₃(C/N)₂T_x

Ti₃C₂T_x is the first synthesized and the most-commonly used MXene material. Compared with other metal-based catalysts, Ti₃(C/N)₂T_x features strong thermal stability, which is ascribed to the M_n+1X_n skeleton formed by carbonic atomic (C) layers and transition metal (Ti) atomic layers.¹⁰¹ Besides, the surface terminations generated during the synthetic process contribute a variety of properties such as hydrophilicity and highly diffusing Lewis acidity to Ti₃(C/N)₂T_x.¹⁰² Therefore, Ti₃(C/N)₂T_x shows promising catalytic performance in hydrogenation,^93,94 dehydrogenation,^41,95 and desulfurization reactions of complex organic compounds, such as biomass and polycyclic organics. In this part, we analyze the role of the unique structure of Ti₃(C/N)₂T_x in improving the activity of hydrogenation, dehydrogenation, and other reactions. Moreover, the influence of termination groups and carbide-metal-termination bonds of Ti₃(C/N)₂T_x on the catalytic reactions is discussed.

3.1.1 Hydrogenation. Hydrogenation is the key step in upgrading various lignocellulosic platform chemicals like levulinic acid, furfural, phenols, and guaiacol.^103,104 However, it is still challenging to control the product selectivity in the hydrogenation of biomass-derived compounds, because these reaction pathways are quite complex and accompanied by various side reactions.¹⁰⁴ Naguib et al. first applied Ti-based MXenes as catalysts to selectively hydrogenate furfural (a substance derived from hemicelluloses).⁹³ Under the catalysis of Ti₃C₂T_x with H₂ as the reducing agent, the main product, furfuryl alcohol, was produced at a rate of 145 mmol_furfural g⁻¹ h⁻¹ and a selectivity of 52%, which was higher than that of TiO₂-based materials (rutile, anatase, and P25) under the same conditions. The reason was that the metal–oxygen site pair of MXenes played an important role. On this peculiar pair of MXenes, heterolytic activation of hydrogen occurred, and then the selective hydrogenation of the carbonyl bond proceeded via the addition reaction of the proton and hydride to the C and O atoms. Moreover, the Ti₃CNT_x catalyst showed similar activity (126 mmol_furfural g⁻¹ h⁻¹) and selectivity (49%) for furfuryl alcohol, but higher stability compared with Ti₃C₂T_x. The better performance of Ti₃CNT_x against deactivation was ascribed to the lower coverage reaction products caused by less diverse termination groups on its surface, indicated by a single Ti₃CNT_x (0002) peak at around 8° in XRD corresponding to a c lattice parameter (c-LP) of 21.9 Å (Fig. 6a). In contrast, the deactivation of Ti₃C₂T_x came from multiple intercalation states due to the reaction occurring on the surface termination groups of MXenes, which was proved by the replacement of the (0002) peak in pristine Ti₃C₂T_x (2θ of 8.96°) by three new peaks (3.94°, 5.56°, and 9.24°), as shown in Fig. 6b.⁹³

Fig. 6 The applications of Ti₃C₂T_x in hydrogenation. (a) Ti₃CNT_x before and after hydrogenation. (b) Ti₃C₂T_x before and after hydrogenation. The (0002) peak areas are presented in the insets, and the blue diamonds represent the peak position of anatase [PDF# 02-0406]. Reproduced with permission.⁹³ Copyright 2020, Wiley-VCH. (c) The conversion mechanism of guaiacol, anisole, and phenol. Reproduced with permission.⁹⁴ Copyright 2020, Elsevier.

Besides hydrogenation, upgrading biomass-derived compounds by removing the oxygen groups is equally desirable. A Ti₃C₂T_x nano-sheet was a type of unique catalyst with superior performance in hydrodeoxygenation (HDO).⁹⁴ With this catalyst, guaiacol, a lignin-derived compound, was converted to phenol and methylanisole at a selectivity of 61% and 17%, respectively, at 350 °C and 5.0 MPa. Furthermore, Ti₃C₂T_x served as a bifunctional catalyst simultaneously promoting the demethylation (DME) reaction of anisole to produce phenol (Fig. 6c). In this process, the main products were phenol and methylanisole with similar selectivities of 45% and 38%, respectively. In comparison with metal carbide (Re, Ru, Ni, Fe, and Cu) catalysts, the phenol selectivity of Ti₃C₂T_x was the highest, which can be ascribed to the surface termination groups of MXenes.¹⁰⁵ To be specific, the surface termination groups contained weak basic sites and acid sites. The acid sites were derived from the F termination and Ti element, on which methylanisole has been obtained by bi-molecular transalkylation. The C–Ti–O structure provided basic sites, where the direct demethoxylation (DMO) of guaiacol was expected to occur.

3.1.2 Dehydrogenation. Dehydrogenation plays an important role in the refinery cracking process where saturated alkanes are turned into olefins and aromatic compounds.^106,107 However, during this process, a high selectivity and yield only can be obtained at high temperatures (∼620 °C); hence, excess steam must be provided simultaneously to alleviate catalyst coking.^106,107 To surmount this obstacle, carbon-based catalysts, like graphene and carbon nanotubes, were applied, and Ti₃C₂T_x was also explored to catalyze the direct dehydrogenation reaction of ethylbenzene (EB).¹⁰⁶ In this routine, by using Ti₃C₂T_x as the catalyst, styrene (ST), an important and greatly demanded industrial monomer, was obtained with a high reactivity of 92 μmol m⁻² h⁻¹ and a high selectivity of 97.5% at a relatively low temperature (550 °C). In contrast, the reactivity of the analogous graphene, nanodiamond, and TiC-derived carbon (TiC-CDC) was 12, 7, and 0.87 μmol m⁻² h⁻¹ respectively, much lower than that of Ti₃C₂T_x (Fig. 7a). The reason why the MXenes boosted the dehydrogenation reaction could be ascribed to the C–Ti–O bonds, the most abundant types among multiple chemical environments of O elements (Ti–O, C–Ti–O, C–Ti–OH, and Al–O). On C–Ti–O bonds, the energy barrier of the dehydrogenation process was significantly reduced. According to first-principles calculations, the first C–H bond of EB was activated with an energy barrier of 0.66 eV, and the second hydrogen abstraction barrier was 0.46 eV. Moreover, Ti₃C₂T_x can catalyze the EB conversion at a stable rate for ca. 40 h without obvious deactivation (Fig. 7b). In contrast, commercial iron oxide suffered from significant coke deposition for the same period.

Fig. 7 The dehydrogenation and desulfurization reactions with Ti₃C₂T_x. (a) Ethylbenzene dehydrogenation reactivity with Ti₃C₂T_x MXene, nanodiamond, graphene, and TiC-derived carbon (TiC-CDC). (b) Long-term stability of Ti₃C₂T_x MXene in the dehydrogenation of ethylbenzene. Reproduced with permission.⁹⁵ Copyright 2018, American Chemical Society. (c) DBT conversion with different catalysts. Reproduced with permission.⁹⁶ Copyright 2021, Elsevier. (d) DBT conversion rates with h-TC/TAC-x-y prepared at different alkali concentrations and etching temperatures, where x is the concentration of the KOH etchant, and y is the etching temperature. Reproduced with permission.⁴⁴ Copyright 2022, Elsevier.

Recently, the hydrogen-release reaction is another dehydrogenation reaction that has attracted emerging attention.¹⁰⁸ Hou et al. prepared rich oxygen-covered Ti₃C₂T_x to boost the catalytic activity of HCOOH dehydrogenation.⁴¹ In this reaction, HCOOH was finally converted into CO₂ with the liberation of hydrogen at a space-time-yield (STY) of 365 mmol⁻¹ g⁻¹ h⁻¹ and 100% selectivity. According to DFT calculations, the rich-oxygen-covered surface of Ti₃C₂T_x acted as the active site. On these sites, HCOOH was adsorbed with an adsorption energy of −3.05 eV, followed by spontaneous dissociation to generate HCOO* and H* with an energy barrier (E_a) of only 0.25 eV. Then the dissociation of HCOO* occurred over Ti₃C₂T_x with an E_a of only 0.61 eV, and CO₂ and H₂ were desorbed from Ti₃C₂T_x with the energies of −0.16 and −0.06 eV, respectively. In a word, the remarkable catalytic performance of Ti₃C₂T_x was due to the low reaction energy barrier of the HCOO*-to-CO₂* step and low desorption energies of CO₂ and H₂.

3.1.3 Desulfurization. Oxidative desulfurization (ODS) is important to reduce SO_x emissions from petroleum fuel combustion processes,¹⁰⁹ but it often has low conversion and requires harsh conditions.¹¹⁰ Transition metal carbides (W, Mo, V, Zr, and Ti) are commonly used as the catalysts for this reaction. For example, Ti₃C₂ was adopted to remove dibenzothiophene (DBT) and its derivative sulfides.⁹⁶ With a catalyst concentration of only 0.25 mg mL⁻¹, the DBT conversion reached 95.7% (Fig. 7c). By contrast, the conversion with TiO₂/graphite was merely 5.6%, and that with the MAX phase Ti₃AlC₂ was only 17.2%. Furthermore, when Ti₃C₂ was employed for the conversion of real diesel, the total sulfur concentration could be reduced from 186 to 24 ppm, and the total sulfur removal reached 87%. This was because the coordinated unsaturated metal Ti sites in the Ti₃C₂ MXene were easily oxidized by molecular oxygen to form TiO₂ clusters. The defective TiO₂ clusters generated dynamically on the surface of MXenes were highly active reaction sites where O₂ was converted into O₂²⁻ and ˙O₂⁻, and further converted into RC(OOO)˙ and ˙OH radicals.

Additionally, a kind of hydroxylated Ti₃C₂/Ti₃AlC₂ (h-TC/TAC) catalyst was prepared for the aerobic desulfurization of DBT.⁴⁴ The h-TC/TAC prepared in a 5 mol L⁻¹ KOH etchant at 120 °C for 24 h exhibited a desulfurization rate of up to 99% (Fig. 7d) and can be recycled eight times without significant deactivation. In this reaction, the oxygen species on the catalyst surface including Ti–O and Ti–OH could be considered as possible active species for ODS, on which the molecular oxygen was converted into ˙OH and ˙O₂⁻ continuously. Therefore, DBT can be easily absorbed on the surface of the catalysts and further oxidized by these two radicals.

3.2 Mo₂C(T_x)

3D molybdenum carbide Mo₂C has been widely applied in CO₂ reduction,^111–114 propane activation,¹¹⁵ and HDO reactions, due to its platinum-like catalytic performance.^{39,40,116,117} More recently, 2D molybdenum carbide (2D-Mo₂C or Mo₂CT_x), has shown promising performance in CO₂ reduction^51,98 and the water-gas shift (WGS) reaction.¹⁰⁰ In this part, we summarize the effect of the pores in layered structures on accelerating the adsorption process and try to clarify the effect of surface termination groups on catalyst activity and stability.

3.2.1 CO₂ reduction (dry reforming and CO₂ hydrogenation). The reduction of CO₂ to valuable chemicals and fuels to mitigate the greenhouse effect has received wide attention.^118,119 However, there are two major obstacles that limit its development, i.e., the CO₂ adsorption energy barrier^120,121 and difficulties in precisely controlling the selectivity of products.¹²² Recently, Mo₂C(T_x) was applied in the CO₂ reduction reaction as a new type of catalyst.⁵¹ For instance, a multilayered hexagonal 2D-Mo₂C material with only Mo-terminated basal planes was synthesized by treating Mo₂CT_x at 500 °C for 2 h under 100% H₂. This material featured high selectivity and activity in reverse water–gas shift (RWGS). To be precise, under the catalysis of 2D-Mo₂C, the formation rate of CO reached 475 mg h⁻¹ g_cat⁻¹, and the selectivity of CO was 94%. However, the CO formation rates with β-Mo₂C and Mo₂CT_x−700, which sintered into the 3D structure, are only 17% and 33% of that with 2D-Mo₂C, respectively (Fig. 8a). According to CO temperature-programmed desorption (TPD) experiments, on the one hand, this high catalytic activity stemmed from mass transfer effects. In 2D-Mo₂C, reactive gas can be adsorbed both on the surface and inside the 2D pores of this multilayered material, indicated by a broad CO desorption peak at 24 °C (Fig. 8b). However, the adsorption only occurred on the surface of β-Mo₂C and Mo₂CT_x−700 due to their bulk structures. On the other hand, the high activity was attributed to the high exposure of Mo sites on 2D-Mo₂C without the blocking of the termination groups (–T_x).

Fig. 8 The thermal catalysis with Mo₂CT_x. (a) Intrinsic formation rates of β-Mo₂C, Mo₂CT_x, 2D-Mo₂C, and Mo₂CT_x-700. (b) CO temperature-programmed desorption (TPD) of 2D-Mo₂C and β-Mo₂C. Reproduced with permission.⁵¹ Copyright 2021, Springer Nature. (c) Catalytic performance of 2D-Mo₂C/SiO₂ (90–150 min) and 2D-Mo₂CO_x/SiO₂ (240–300 min and from 390 min after the CO₂ regeneration step) in DRM. (d) Correlation of the Mo oxidation state (determined by Mo K-edge XANES), and the catalytic activity of 2D-Mo₂C/SiO₂, 2D-Mo₂CO_x/SiO_2-TOS10, and 2D-Mo₂C/SiO_2-TOS100. Reproduced with permission.⁹⁸ Copyright 2020, Springer Nature. (e) WGS catalytic activity of Mo₂CT_x, Mo₂CT_x-500, and a reference β-Mo₂C. (f) In situ structural stability of Mo₂CT_x studied by TGA H₂-TPR. Reproduced with permission.¹⁰⁰ Copyright 2019, American Chemical Society.

Moreover, two-dimensional molybdenum oxycarbide supported on silica (2D-Mo₂CO_x/SiO₂) was found to have great activity and selectivity for the dry reforming of methane (DRM).⁹⁸ The synthesis of 2D-Mo₂CO_x/SiO₂ was divided into two steps. The first step was treating 2D-Mo₂CT_x/SiO₂ under H₂ (750 °C, 0.5 h) to prepare 2D-Mo₂C/SiO₂, and the second step was exposing 2D-Mo₂CT_x/SiO₂ to a CO₂ atmosphere to prepare 2D-Mo₂CO_x/SiO₂. Compared with 2D-Mo₂C/SiO₂ showing negligible DRM activity, 2D-Mo₂CO_x/SiO₂ was highly active with a methane conversion rate of 0.42 mol_(CH4) mol_(Mo)⁻¹ s⁻¹ after a time on stream (TOS) of 10 min at 80% CH₄ conversion (Fig. 8c). The improved catalytic activity of 2D-Mo₂CO_x/SiO₂ can be attributed to the suitable oxidation state of Mo in the 2D-nanosheet catalyst. XANES indicated that if the oxidation of 2D-Mo₂C/SiO₂ was carried out under DRM conditions (730 °C, CH₄ : CO₂ = 1 : 1, space velocity (SV) = 3 × 10⁴ L g_Mo⁻¹ h⁻¹, and contact time of 0.1 ms g_Mo mL⁻¹) for a TOS of 10 min (2D-Mo₂COh_x/SiO_2-TOS10), the Mo oxidation state of 2D-Mo₂CO_x/SiO₂ was +4, which was highly active in DRM (Fig. 8d). However, if the oxidation of 2D-Mo₂C/SiO₂ was conducted at 800 °C in a CO₂ atmosphere, the surface of 2D-Mo₂CO_x/SiO₂ was covered by oxygen and the average Mo oxidation state was +5.5, which was inactive for DRM. Moreover, DFT calculations further explained why the exact oxidation state resulted in the higher activity, i.e., the suitable oxygen coverage played a significant role in reducing the binding energy of C* and O* species and lowered the energy barriers of the C–O coupling in the CH₄ cleavage process, which was the rate-limiting step in DRM.

3.2.2 Water–gas shift reaction. The water–gas shift (WGS) reaction is a significant step in many industrial processes, including ammonia synthesis, methanol synthesis, and hydrogen production. However, the current industrial catalysts are unstable and normally require lengthy activation steps.^123–125 In a recent study, Mo₂CT_x was applied in WGS with high activity and stability.¹⁰⁰ At around 520 °C, Mo₂CT_x reached the maximal activity with a CO consumption rate of ca. 100 μmol_(CO) g_(Mo)⁻¹ s⁻¹ and a selectivity of 99%. However, under similar conditions, the activity of Mo₂CT_x-500 (reducing Mo₂CT_x at 500 °C in 20% H₂/N₂) was much lower than that of Mo₂CT_x, and the reference material β-Mo₂C was completely inactive for this reaction (Fig. 8e). This phenomenon was attributed to the mass transfer limitation. In Mo₂CT_x-500, its interlayer distance decreased, which resulted in the reduction of reactant transfer capacity. Meanwhile in β-Mo₂C, the mass transport process was hindered by its bulk structure. In the H₂ TPR study, only interlayer water was desorbed from Mo₂CT_x below 350 °C, and it did not undergo any structural change. When the temperature reached 500 °C, the interlayer spacing of Mo₂CT_x decreased obviously due to the violent de-functionalization and the removal of the intercalated water, but it can still maintain the integrity of the structure. When the temperature reached 600 °C, two-dimensional Mo₂CT_x sheets started sintering to a bulk carbide phase (Fig. 8f).

3.3 V₂CT_x-derived catalysts

V₂CT_x is another commonly reported and used MXene material after Ti₃CT_x. Thakur et al. employed V₂CT_x in methane dehydroaromatization (MDA) to directly convert methane (CH₄) into benzene (C₆H₆) at 700 °C.⁹⁹ The catalyst showed a state-of-the-art CH₄ conversion of 11.8% with a C₆H₆ formation rate of 1.9 mmol g_cat⁻¹ h⁻¹ at 700 °C (Fig. 9a). Conversely, V₂AlC exhibited no reactivity in the MDA reaction. The good performance, on the one hand, was attributed to the interlamellar space, which provided the confinement for the oligomerization and cyclization of intermediates (C2 species) to then form benzene. On the other hand, the Lewis and Brønsted sites originating from surface termination groups made a significant contribution to the oligomerization reaction of *C₂H₃ (intermediates of MDA) to form benzene (Fig. 9b). However, at this temperature, the MXene structure was converted into amorphous non-MXene phases. Moreover, after the long-term reaction, some solid amorphous and graphitic carbon grew on the surface of V₂CT_x dervied catalysts. (a) CH4 conversion and (Fig. 9c), limiting the accessibility of CH₄ inside the channels, resulting in the deactivation of this catalyst.

Fig. 9 The thermal catalysis with V₂CT_x-dervied catalysts. (a) CH₄ conversion and C₆H₆ formation rates with V₂CT_x. (b) Schematic illustration of forming C₆H₆ on V₂CT_x-dervied catalysts. (c) The relationship between the relative formation rate of D (1351 cm⁻¹) and G (1592 cm⁻¹) bands and TOS. Reproduced with permission.⁹⁹ Copyright 2020, Wiley-VCH. (d) CH₄ conversion and H₂/CO ratio with VC, V₂AlC, V₂O₃–V₈C₇/m-V₂CT_x MXene, and 5Ni/ZSM-5. (e) Proposed DRM mechanism using the m-V₂CT_x MXene-derived catalyst. (f) CH₄ and CO₂ conversion activity and H₂/CO ratio with V₂O₃–V₈C₇/m-V₂CT_x of surface termination groups for 96 h. Reproduced with permission.⁹⁷ Copyright 2020, American Chemical Society.

Moreover, multilayered vanadium carbide (m-V₂CT_x) can also serve as the precursor for a robust oxy-carbide catalyst in DRM.⁹⁷ This material showed an attractive activity in DRM, converting about 78% of CH₄ with a rate of 153.4 mmol_CH4 mol_V⁻¹ min⁻¹, which was comparable to the CH₄ conversion rate of Ni-based catalysts and was about four times higher than that of its bulk counterparts V₂AlC MAX phase and commercial VC (Fig. 9d). In this DRM process, V₂O₃ and V₈C₇ nanocrystals were generated in situ both on the surface and layered structure of V₂CT_x, resulting in maximizing the utilization of active and selective V sites on the new V₂O–V₈C₇/m-V₂CT_x. The detailed process was revealed by isotopic labeling experiments with ¹³CO₂ and ¹³CH₄. Initially, when heated to 800 °C under an inert gas, m-V₂CT_x was partially oxidized to V₂O₃, while some of the MXene transformed into V₈C₇ particles. V₈C₇/m-V₂CT_x was oxidized by CO₂ to form V₂O₃/m-V₂CT_x and CO. Meanwhile, recarburization of V₂O₃/m-V₂CT_x with CH₄ resulted in the formation of V₈C₇/m-V₂CT_x CO and H₂ (Fig. 9e). Therefore, the competition between the oxidation reaction and the carburization reaction resulted in a more stable material with the CH₄ and CO₂ conversion rates of 1.0 mmol_(CH4) g_cat⁻¹ min⁻¹ and 1.3 mmol_(CO2) g_cat⁻¹ min⁻¹ respectively even after 96 h on stream (Fig. 9f).⁹⁷

3.4 Influence of surface terminations

Many significant properties of MXenes such as hydrophilicity, conductivity,⁵⁸ and stability can be tuned by controlling the number and variety of surface termination groups T_x. Similarly, the activity and selectivity of MXenes in thermal catalytic reactions can also be affected by changing T_x groups. This is attributed to the influence of T_x groups on adjacent transition metal atoms and surface properties. Therefore, in this part, we sum up the methods to vary the types and coverage of the surface termination groups and further summarize the role of T_x from three aspects: (i) influencing the exposure of the transition metals, (ii) affecting the oxidation states of the adjacent transition metal atoms, and (iii) altering the acid–base properties (Fig. 10).

Fig. 10 Schematic diagram of the influence of surface termination groups.

The variation of surface terminations can be realized by heat treatment, in which the temperature and atmosphere are two significant factors. Under a reductive environment, surface termination groups are removed. In dilute H₂, only partial T_x groups can be removed from Mo₂CT_x even when the temperature rose to 500 °C.⁵¹ Specifically, when 5% H₂/Ar acted as the reducing gas, hydroxyl and part of fluorine were reduced from the Mo₂CT_x surface at 280 °C, and subsequently, oxo groups and most of the fluorine were removed at 500 °C. However, if pure H₂ was used, the removal of partial T_x occurred at a relatively lower temperature at 175 °C, while at 500 °C all the surface termination groups were totally defunctionalized from the Mo₂CT_x surface. The hydrogen treatment process of V₂CT_x was also investigated, in which the removal of T_x groups occurred above 300 °C, and VO_x species on its surface were reduced simultaneously.¹²⁶ Furthermore, the termination variation of V₂CT_x in inert and oxidative environments was also researched. The result was that N₂-treated V₂CT_x was extremely stable even up to 600 °C, and only minor oxidation occurred on its surface. Under CO₂, the material was oxidized above 300 °C, leading to the surface covering a larger percentage of VO_x species.

More transition metal atoms will be exposed as T_x groups are removed. In the comparison of Mo₂CT_x and 2D-Mo₂C, 2D-Mo₂C had much more exposed Mo sites (Fig. 11a), indicated by the high CO uptake up to 41.1 μmol g⁻¹. These exposed Mo sites acted as active sites to facilitate CO₂ adsorption and H₂ dissociation, boosting the activity of 2D-Mo₂C in RWGS.⁵¹ Mo₂CT_x is an excellent catalyst for the WGS reaction,¹⁰⁰ but it is inactive in the RWGS reaction. However, 2D-Mo₂C, showing high activity in RWGS, has a poor catalytic performance in the WGS reaction. XANES data revealed that the Mo got oxidized under the WGS conditions, reaching an oxidation state close to Mo₂CT_x. The lower activity was then due to the mass transport limitations caused by the decreased interlayer distance.⁵¹

Fig. 11 Influence of surface terminations. (a) Schematic of the exposure of Mo sites in the formation of 2D-Mo₂C from Mo₂CT_x. Reproduced with permission.⁵¹ Copyright 2021, Springer Nature. (b) Oxidation state of molybdenum in reference and synthesized materials determined from XANES spectra. Reproduced with permission.⁹⁸ Copyright 2020, Springer Nature. (c) The mechanism schematic diagram of producing phenol and methylanisole from guaiacol with the acid-basic pair. Reproduced with permission.⁹⁴ Copyright 2020, Elsevier.

Meanwhile, the surface termination groups affect the oxidation states of the transition metal. Taking Mo₂CT_x and 2D-Mo₂C (without T_x) as an example, the original Mo₂CT_x has a Mo oxidation state of ca. +3.9. After the surface termination groups were released, the oxidation states of Mo decreased, leading to a Mo oxidation state of +0.5 in 2D-Mo₂C.¹⁰⁰ The oxidation states of the transition metal may determine the activity of the material in the catalysis. To be specific, if the oxygen coverage of 2D-Mo₂CO_x/SiO₂ was either too low or too high corresponding to the Mo oxidation states of ca. +0.2 and +5.5 (Fig. 11b), it was inactive in the DRM reaction. In contrast, 2D-Mo₂CO_x/SiO₂ with two-thirds oxygen coverage corresponding to a +4 Mo oxidation state was a kind of excellent catalyst in DRM, as evidenced by the DFT calculation.⁹⁸

Surface termination groups also play an important role in forming acid-basic sites and active sites, which can promote the activity of reaction. Compared with β-Mo₂C, Brønsted or Lewis acid sites were observed on Mo₂CT_x, which could promote the dehydration of methanol to produce dimethyl ether (DME). Similar acidity was observed on 2D-Mo₂C, where T_x groups were completely removed.⁵¹ Furthermore, the fluoro groups can strongly bond with transition metals in Ti₃C₂T_x to form active Brønsted acid sites, while the surface oxygen groups were identified to be active basic sites (Fig. 11c). On the surface covered with these acid-basic pairs, furfural hydrogenation and guaiacol HDO were greatly promoted.^93,94

3.5 Active sites

Active sites are typically the exposed transition metals or the unique structure of the metals and surface termination groups of MXenes. In contrast with termination groups, the regulation of active sites is a more direct way to improve the efficiency of thermal catalysis, because these sites will promote reactants to form free radical particles or reduce the reaction energy barriers.¹²⁷

As for the function in facilitating the generation of free radical particles, the desulfurization reaction catalyzed by h-TC/TAC is a good example.^44,96 In this reaction, the –O and –OH termination groups produced in the alkaline selective etching can interact strongly with Ti to form Ti–O and Ti–OH structures, and in turn, these unique structures act as active centers to accelerate the oxygen molecular transformation into radicals. In situ electron spin resonance (ESR) spectroscopy revealed that the oxygen molecules were likely to be converted into ˙OH and ˙O₂⁻ radicals on these Ti–O and Ti–OH structures. Then these oxygen radicals would interact with DBT to further generate DBTO₂ (Fig. 12a).⁴⁴ In addition, Ti metal active sites in Ti₃C₂ MXenes also promoted the production of radicals in the thermal decomposition of ammonium perchlorate-based molecule perovskite ((H₂dabco)[NH₄(ClO₄)₃], DAP-4).¹²⁸ In this case, the protons were transferred from H₂dabco²⁺ and NH⁴⁺ to ClO₄⁻ to generate dabco, NH₃, and HClO₄, respectively. Among these products, the oxygen atom in HClO₄ was further converted into the superoxide radical anion (˙O₂⁻) with the assistance of Ti sites (Fig. 12b). Under the action of ˙O₂⁻, the cage structure of DAP-4 composed of NH⁴⁺, ClO⁴⁻, and H₂dabco²⁺ was destroyed completely which made the combustion reaction occur at a relatively low temperature (341.2 °C) compared with the reaction temperature of thermal decomposition of DAP-4 without any catalysts at 385 °C.

Fig. 12 The roles of active sites. (a) The DBT activation process under the action of ˙O₂⁻ and ˙OH radicals. Reproduced with permission.⁴⁴ Copyright 2020, Elsevier. (b) The thermal decomposition of DAP-4 with the assistance of ˙O₂⁻. Reproduced with permission.¹²⁸ Copyright 2020, Elsevier. (c) The reaction pathway and important structures of ethylbenzene dehydrogenation on Ti₃C₂O₂. Ti is shown in light gray, C in dark gray, O in red, and H in white. Reproduced with permission.⁹⁵ Copyright 2018, American Chemical Society.

The active sites from MXene materials can decrease the energy barrier of the rate-limiting step. For example, José and co-workers used DFT to calculate the energy barrier of ammonia (NH₃) production catalyzed by 18 different MXenes including nine carbides and nine nitrides of which transition metals ranged from groups IV to group VI.⁴³ The result showed that with the assistance of hollow metal sites on the MXene (0001) surface, the dissociation of molecular nitrogen, the rate-limiting step for ammonia production, was greatly promoted due to its low energy barriers under 1 eV. W₂N was the most favorable type of MXene for this reaction with a dissociation energy barrier of only 0.28 eV, which was lower than that of Ru metal and a cesium-promoted ruthenium particle supported on MgO. Likewise, hydrogen abstraction, the rate-limiting step of the dehydrogenation reaction, was facilitated as well due to the reduced energy barriers. To be precise, on the oxygen-covered surface, the dissociation energies of the two hydrogen atoms in HCOOH were only 0.25 and 0.61 eV, respectively. Besides, the energies for breaking the two C–H bonds in EB were merely 0.66 and 0.46 eV, respectively (Fig. 12c).

Furthermore, the reduction of energy barriers can be realized by decreasing the adsorption energy, which was suggested by calculating the adsorption and dissociation energy barriers of water on a set of 18 M₂X MXenes (M = Ti, Zr, Hf, V, Nb, Ta, Cr, Mo, and W, while X = C or N).¹²⁹ It was shown that all the studied MXenes exothermically adsorbed water on their corresponding metal sites with the adsorption energies ranging from −1.43 to −2.94 eV, which was much lower compared with the ternary catalyst with Zn and Al atoms doped on Cu surfaces.

4. MXenes as supports

MXenes can not only serve as catalysts to directly provide active sites via their robust M_n+1X_n frameworks and surface termination groups but also form more active heterogeneous catalysts by supporting transition metals (Table 2). The advantages of adopting MXenes as catalyst supports include their tunable surface properties¹³⁰ and outstanding stability,¹³¹ which can offer the anchoring sites for transition metals and improve their anti-sintering ability.^132,133 Nanoparticles (NPs), nanoclusters, or single atoms can be formed on MXene supports according to the different sizes of the transition metals. In this section, we thoroughly summarize the reactions catalyzed by MXene-supported catalysts (Fig. 13) and explain the unique advantages of MXenes as supports. Furthermore, the metal–support interactions between transition metals and MXenes are explicated.

Table 2 The catalytic performance of MXene-supported catalysts

Catalyst	Reaction	Main reactant	T (°C)	P (MPa)	Conv.^a (%)	Target product	Sel.^b (%)	Ref.
a Conversion refers to the conversion of the main reactant. b Selectivity refers to the selectivity of the target product.
Pd50–Ru50/Ti3C2Tx	CO2 hydrogenation	CO2	150	1.0	78	CH3OH	76	132
Co–Ti3C2Tx	CO2 hydrogenation	CO2	400	0.1	15.7	CO	86.4	134
Co–Ti3C2Tx-NH3	CO2 hydrogenation	CO2	400	0.1	45.1	CH4	79.6	134
Pt1/Ti3−xC2Tx	CO2 functionalization	CO2	140	0.1	100	Formamides	100	135
Cu/Mo2CTx/SiO2	CO2 hydrogenation	CO2	230	2.5	3.2	CH3OH	37	122
Pt/Ti3C2Tx	Light alkane dehydrogenation	Propane	550	0.1	15	Propylene	∼95	136
Pt/Nb2CTx	Light alkane dehydrogenation	Propane	550	0.1	15	Propylene	∼90	136
Pt/Nb2CTx	Water–gas shift reaction	CO	300	0.1	—	H2	—	137
Pt/Mo2TiC2Tx	Non-oxidative coupling of methane	CH4	750	0.1	7	C2 products	>98	138

---

Fig. 13 Catalysts formed by the interactions between MXene supports and metals and their applications in thermal catalysis.

4.1 Transition metal nanoparticles (NPs)

Transition metal nanoparticles (TMNPs), the state between the bulk solid and molecular level, are the common loading form on MXenes,¹³⁹ due to their higher dispersion in solvent and 3D rotational freedom.^140,141 Combined with the strong catalytic activity from transition metals and the stability from MXenes,^136,137 MXene-supported catalysts can be obtained through some relatively simple preparation methods such as impregnation. In this part, we summarize the common way to load TMNPs on MXenes and further elaborate the application of these catalysts in thermal catalysis.

4.1.1 Synthesis. The most common way to generate MXene-supported TMNPs is impregnation. In this method, the catalyst supports are first immersed or soaked in a metal salt solution and then they are further activated to generate their corresponding active forms.¹⁴² Heat treatment is a commonly used activation method. For example, Ma et al. used this method to prepare Co–Ti₃C₂T_x MXene.¹³⁴ Ti₃C₂T_x was soaked in cobalt nitrate hexahydrate, and the product obtained from the solution was thermally treated under N₂ and Ar atmospheres. The results showed that Co was deposited on MXenes in the form of rod-shaped Co₃O₄ with a crystal size of ca. 10.8 nm. Besides, nanocomposites can be activated via self-reduction reactions. For instance, Pt/Ti₃C₂T_x can be prepared by dispersing it in a solution of H₂PtCl₆,¹⁴³ and the corresponding product was reduced with ammonia borane (NH₃BH₃). The results showed that the Pt TMNPs were clearly distributed on the Ti₃C₂T_x surface with an average size of 2.2 nm. Furthermore, a Ti₃C₂(OH_xF_1−x)₂@AuNP composite was prepared by a similar process,⁴⁷ in which the MXenes were added into HAuCl₄ with the reducing agent of NaBH₄. Recently, this impregnation route was accelerated with the assistance of microwaves.¹³² Under the microwaves, PdCl₂, RuCl₃·xH₂O salts, and NaBH₄ were added to the Ti₃C₂T_x colloidal solution in turn. In this case, metallic Pd and Ru nucleation seeds were formed on Ti₃C₂T_x, which favored the formation of Pd₅₀–Ru₅₀ alloy particles. Eventually, the individual Pd₅₀–Ru₅₀ particles displayed a highly agglomeration-free homogeneous dispersion on Ti₃C₂T_x with an average diameter of 3.2 nm.

4.1.2 Applications in thermal catalysis. MXenes especially Ti₃C₂T_x usually act as supports to improve the activity and stability of the catalysts in CO₂ reduction and dehydrogenation of hydrogen carriers. Pd₅₀–Ru₅₀/Ti₃C₂T_x has been used in the CO₂ hydrogenation process in which NaBH₄ served as a H₂ donor and ethylene glycol (EG) solution acted as a capture agent.¹³² Under the optimal reaction conditions (at 120 °C for 12 h in a mixed solution of EG and water with a volume ratio of 18 : 2, and hydrogen was provided by 1 mol NaBH₄), the Pd₅₀–Ru₅₀/Ti₃C₂T_x catalyst produced a higher CO₂ conversion efficiency of 78% with a CH₃OH yield of 76%. In contrast, Ti₃C₂T_x and Pd₅₀–Ru₅₀ displayed a CH₃OH yield of 23% and 52%, and a total CO₂ conversion of 28% and 57%, respectively (Fig. 14a). The superior catalytic effect stemmed from the synergy of the Pd₅₀–Ru₅₀ alloy and Ti₃C₂T_x, in which the Ti–O basic pair on Ti₃C₂T_x facilitated the adsorption of CO₂. Meanwhile, the interface between Pd₅₀–Ru₅₀ and MXenes was suitable for adsorbing H₂ to generate H* atoms. After the adsorption process, the reactive H* atoms diffused to the vicinity of adsorbed CO₂ molecules to generate HCOO* followed by the formation of H₂COO*. The H₂COO* would then react with excessive H* atoms to form H₂CO* associated with water molecules, resulting in methanol formation (Fig. 14b). Moreover, it has also been reported that the Pd₅₀–Ru₅₀/Ti₃C₂T_x catalyst showed remarkable chemical stability even after prolonged use for CO₂ hydrogenation.

Fig. 14 The applications of MXene-supported nanoparticle catalysts. (a) The CO₂ hydrogenation with MXene, Pd₅₀–Ru₅₀, and Pd₅₀–Ru₅₀/MXene. (b) Schematic illustration of the hydrogenation of CO₂ to methanol on the Pd₅₀–Ru₅₀/Ti₃C₂T_x catalyst. Reproduced with permission.¹³² Copyright 2021, Elsevier. (c) The reaction pathway over cobalt NPs supported on MXenes. Reproduced with permission.¹³⁴ Copyright 2021, Wiley-VCH. (d) Ti 2p XPS spectra of 1 wt% Pt-loaded catalysts. Reproduced with permission.¹⁴⁴ Copyright 2020, IOP.

Co–Ti₃C₂T_x was another catalyst applied in catalyzing CO₂ reduction, and this catalyst realized tunable selectivity during CO₂ hydrogenation by modifying the MXene support.¹³⁴ In Co–Ti₃C₂T_x, CO was the main product with a reaction rate of 10.6 μmol g_cat⁻¹ s⁻¹ and a selectivity of up to 86.4%, which was about three times higher than the CO selectivity of Co/TiO₂ and Co/TiC. However, if N was doped on Co–Ti₃C₂T_x through NH₃ treatment, denoted as Co–Ti₃C₂T_x–NH₃, CH₄ would be the more accessible product with a formation rate of 30.4 μmol g_cat⁻¹ s⁻¹. Co–Ti₃C₂T_x–NH₃ exhibited a CH₄ selectivity of 79.6% and a CO selectivity of 20.4% (Fig. 14c). This change stemmed from the enhanced reducibility of the catalyst surface. Specifically, after N was doped on MXenes by partially substituting C, MXenes would become unstable and TiO₂ emerged on MXenes. The produced TiO₂ strongly interacted with Co nanoparticles and the strong interaction made the reducibility of Co nanoparticles at the interface of the TiO₂ nanoparticles increase, resulting in the shift of product selectivity from CO to CH₄.

Modified MXenes can serve as important supports for dehydrogenation reactions occurring on hydrogen carriers. For instance, Pd/PDA–Ti₃C₂T_x was a superior catalyst to catalyze the formic acid dehydrogenation process, wherein the carrier, PDA–Ti₃C₂T_x, was prepared by alkalization of p-phenylenediamine (PDA), and Pd NPs were loaded on Ti₃C₂T_x by impregnation.¹⁴⁵ This catalyst exhibited a good catalytic activity with a turnover frequency (TOF) value of up to 924.4 h⁻¹, which was 16 times higher than that of Pd/zeolite under similar conditions. The authors proposed that the excellent activity was mainly attributed to the improved synergistic effect between PDA-MXenes and Pd NPs. The PDA together with surface termination groups on MXenes had a scavenging effect on protons, which promoted the dissociation process of the O–H bond in formic acid to generate formate intermediates. Meanwhile the Pd TMNPs provided adsorption sites for formate intermediates and generated palladium formate complexes. On these Pd sites, the C–H bonds of formate intermediates were broken, producing Pd–H and CO₂, and the H of Pd–H species can be further released to produce H₂.

Another commonly used modification method was oxidation. For instance, Ti₃C₂T_x oxidized by O₃ can act as carriers to support Pt NPs to boost the effectiveness of the hydrolysis of ammonia borane.¹⁴⁴ Pt supported on untreated Ti₃C₂T_x (Pt/Ti₃C₂T_x) or commercial TiO₂ (Pt/TiO₂) gave a TOF of 39 and 38 min⁻¹, respectively, whereas supporting Pt on ozone-treated Ti₃C₂T_x (Pt/Ti₃C₂T_x–O₃) increased the TOF by seven times to 265 min⁻¹. Although Pt/Ti₃C₂T_x–H₂O obtained via hydrothermal treatment featured a high TOF similar to Pt/Ti₃C₂T_x–O₃, the activation energy of this material was up to 99 kJ mol⁻¹. This superior catalytic performance was due to electronic interactions between the particles and the modified oxide surface. On the MXene surface, electrons tended to transfer from the TiO₂ layer to the supported particle, making the Pt more electron-deficient with a higher oxidation state (Fig. 14d). Therefore, electron-rich ammonia borane was more likely to be adsorbed on the electron-deficient Pt surface, resulting in lower binding energy.

4.2 Transition metal nanoclusters (NCs) or single atoms (SAs)

Unlike catalysts loaded with NPs, NCs and SACs are usually dispersed atomically on carriers with coordinative unsaturation,¹⁴⁶ making these catalysts have the maximum atomic utilization.^{138,147–149} Unfortunately, the large-scale synthesis of catalysts loaded with NCs or SACs remains a challenge due to the natural tendency for metal atoms to diffuse and agglomerate.¹⁵⁰ In contrast, MXenes can provide numerous dispersed sites for uniform deposition of NCs and SACs through their termination groups and vacancy defects. Therefore, the agglomeration of metal atoms can be avoided, making the MXenes loaded with NCs or SAs become a novel kind of efficient catalyst.¹³⁷ In this section, we exemplify the synthesis methods of this kind of catalyst and provide evidence from both experimental and computational levels that MXene-supported single-atom and atom-cluster catalysts feature strong catalytic activity in various reactions.

4.2.1 Synthesis. Through the surface organometallic chemistry (SOMC) approach followed by H₂ treatment, Cu SAs and NCs were loaded on Mo₂CT_x.¹²² Mo₂CT_x/SiO_2-500, the catalyst carrier, was generated by impregnating 2D-Mo₂CT_x on SiO₂, followed by the dehydroxylated process at 500 °C under a N₂ flow. CuMes [Cu_n(mesityl)_n, n = 2, 4, 5] were grafted onto ≡SiOH sites by dispersing Mo₂CT_x/SiO_2-500 in a mixture solution of toluene and [Cu₅(mesityl)₅]. Then the grafted CuMes precursor was decomposed and migrated from the ≡SiOH sites to 2D-Mo₂CT_x by exposing this grafted material in H₂ at 500 °C for 2 h, and the obtained products were denoted as Cu/Mo₂CT_x/SiO_2-2h. Interestingly, the diameter of Cu on the SiO₂ surface was approximately 3 nm. In contrast, the Cu on the Mo₂CT_x nanosheets had a higher dispersion, forming few-atom-small clusters and single Cu atoms. Although this method can obtain MXene-supported catalysts loaded with Cu NCs and SAs efficiently, the synthetic process was relatively complicated. Pt₁/Ti_3−xC₂T_x catalysts can be prepared by a relatively simple simultaneous self-reduction stabilization process with the help of Ti vacancies under milder conditions.¹³⁵ The H₂PtCl₆·6H₂O solution with [PtCl₆]²⁻ complex ions was injected into the MXene suspension under stirring, and the [PtCl₆]²⁻ ions were uniformly adsorbed onto the surface of Ti_3−xC₂T_x flakes. In this case, Pt⁴⁺ in [PtCl₆]²⁻ was slowly reduced by the highly unstable and reductive Ti vacancies, and Pt₁/Ti_3−xC₂T_x was successfully produced. The formed Pt SAs were uniformly located on Ti vacancies of the Ti_3−xC₂T_x nanosheet by substitution instead of just interstitial doping.

4.2.2 Applications in thermal catalysis. Researchers have demonstrated that MXene-supported single-atom or atom-cluster catalysts have high activity for CO₂ reduction. For example, Zhao et al. built a Pt-loaded single-atom catalyst (Pt₁/Ti_3−xC₂T_x) which offered a green route to utilize greenhouse gas via the formylation of amines, silane, and CO₂.¹³⁵ Pt₁/Ti_3−xC₂T_x shows nearly 100% formamides yield, much higher than the yield with Pt NPs. DFT calculations proved that Pt₁/Ti_3−xC₂T_x showed a relatively lower barrier energy in CO₂ insertion into a Pt–H bond (TS1), the aniline reaction (TS2), and the reductive elimination (TS3) process (Fig. 15a). The authors speculated that the partially positively charged Pt single atoms appeared to decrease the larger steric hindrance in the adsorption of aniline, and simultaneously reduce the energy barrier to active silane, CO₂, and aniline, thereby boosting catalytic performance.

Fig. 15 The applications of MXene-supported nanocluster and single-atom catalysts. (a) Calculated energy profiles for Pt₁/Ti_3−xC₂T_x (red line) and Pt NPs (black line). Reproduced with permission.¹³⁵ Copyright 2019, American Chemical Society. (b) Schematic of Cu/Mo₂CT_x/SiO₂ to selectively promote CO₂ methanolization. (c) Comparison of the intrinsic formation rates of CH₃OH and CO with the different catalysts. Reproduced with permission.¹²² Copyright 2021, Springer Nature. (d) CO oxidation on the Ti-anchored Ti₂CO₂ monolayer via the ER mechanism. Reproduced with permission.¹⁵¹ Copyright 2016, Royal Society of Chemistry. (e) Energy profile and optimized structure of the corresponding stationary points for CO oxidation with Co₁/Mo₂CS₂ by the TER mechanism. The units of energies and bond lengths are eV and Å separately. Reproduced with permission.⁴⁹ Copyright 2021, Springer.

Converting CO₂ into methanol is another way to utilize CO₂ efficiently. Zhou et al. engineered Cu/Mo₂CT_x/SiO₂ to selectivity boost CO₂ methanolization (Fig. 15b).¹²² The activity of Cu/Mo₂CT_x/SiO_2-2h was much higher than that of the industrial Cu-based methanol synthesis catalyst Cu–ZnO–Al₂O₃. Besides, the intrinsic methanol formation rate of Cu/Mo₂CT_x/SiO_2-2h was 1.51 g h⁻¹ g_Cu⁻¹, around three times higher than that of Cu/SiO_2-2h. Through prolonging the reduction time to 6 h in H₂, the Cu/Mo₂CT_x/SiO_2-6h catalyst possessed a higher Cu dispersion compared with Cu/Mo₂CT_x/SiO_2-2h, resulting in a higher intrinsic methanol formation rate of 2.49 g h⁻¹ g_Cu⁻¹ (Fig. 15c). The increased catalytic performance originated from the unique interface formed by the interaction between Cu and Mo₂CT_x. In this interface, the Cu atom and the support both participated in the reaction by reducing the energy barriers for successive heterolytic cleavages of H₂ required to form HCOO*, H₂COO*, and H₂COOH* species. Therefore, the formation of the H₂COO* species, the most energy-demanding step in the methanol production pathway, occurred easily. Hence, on Cu/Mo₂CT_x/SiO_2-6h, CH₃OH was more easily obtained under CO₂ hydrogenation conditions.

Researchers also proved that MXene-supported single-atom and atom-cluster catalysts can greatly decrease the energy barriers in CO oxidation. Zhang et al. discovered a Ti-anchored Ti₂CO₂ monolayer as a single-atom catalyst for CO oxidation.¹⁵¹ The calculation result proved that the reaction was more likely to occur under the Eley–Rideal (ER) mechanism with an energy barrier of 0.23 eV, which was lower than the energy barrier of many noble metal catalysts such as Pt/FeO_x. In this reaction, the physically adsorbed CO was inserted into the O–O bond of the pre-adsorbed O₂ to form a carbonate-like intermediate. Subsequently, another CO molecule was absorbed in the vicinity of a carbonate-like intermediate and transformed into CO₂ (Fig. 15d). The evaluation of Ti-anchored MXene catalysts showed that Ti was adsorbed stably on the Ti₂CO₂ substrate with an energy of −7.68 eV. Co₁/Mo₂CS₂ was another efficient catalyst to accelerate the oxidation of CO, because it can lower the energy barriers of the rate-limiting step in the Termolecular Eley–Rideal (TER) mechanism, i.e., the formation of two CO molecules from the dissociation of the OCO–Co–OCO intermediate, with an energy barrier of only 0.67 eV (Fig. 15e).⁴⁹ Moreover, oxygen vacancies played a significant role in CO oxidation. For instance, the oxygen vacancy in Pd-deposited Mo₂CO₂ can stabilize the single Pd atom, making the Pd/O_V–Mo₂CO₂ system remain an excellent mono-dispersed atomic catalyst in the reaction.¹⁵²

4.3 Metal–support interactions between transition metals and MXenes

By combining the unique properties of metals (in the form of NPs, SAs, and NCs) and MXenes, metal-based catalysts supported by MXenes may be promising in thermal catalysis. While the strong and stable connection between metals and MXenes is usually ignored, it is the prerequisite for the stable use of this kind of catalyst, and therefore, more and more studies start to focus on metal–support interactions. Currently, there are three kinds of interactions for MXene-supported catalysts: (i) electrostatic interaction,¹³² (ii) intermetallic interaction,¹³⁸ and (iii) coordination interaction.¹²² Due to these three kinds of interactions, the synergistic effects in catalysis between a transition metal and MXenes can be obtained, and the stability of MXene-supported catalysts is greatly enhanced.¹⁵³ Therefore, in this section, we first summarize the main interactions between transition metals and MXenes. The effect of the interactions on catalyst properties and catalytic performance is further explained.

Among the numerous MXene-supported catalysts loaded with TMNPs, metal–support interactions usually exist in the form of electrostatic attraction force, which typically originates from the impregnation process. Specifically, when MXenes like Pd₅₀–Ru₅₀/Ti₃C₂T_x (Fig. 16a) were immersed in metal salt solutions, positively charged metal ions were attracted by the negatively charged termination groups on MXenes.¹³² After that, the metal ions can be reduced to metallic particles in situ with a reducing agent or heat treatment.¹³² Furthermore, some stronger electrostatic interactions can be obtained on modified MXenes with NPs loaded, such as Pt/Ti₃C₂T_x–O₃.¹⁴⁴ On these surfaces of MXenes, Ti in unsaturated oxide states including Ti²⁺ and Ti³⁺ was transformed into TiO₂. When the TMNPs approached the TiO₂, strong connections between the MXenes and the NPs occurred, due to the electron tending to transfer from TiO₂ to Pt. Besides, the intermetallic phase was another stable interaction, which emerged on Pt/Ti₃C₂T_x and Pt/Nb₂CT_x treated with H₂.¹³⁶ Under a hydrogen atmosphere, fully exposed Ti or Nb strongly interacted with Pt to form a highly regular intermetallic compound (IMC), Pt₃Ti, via reactive metal–support interactions (RMSI) (Fig. 16b). This compound was composed of periodic hexagonal arrays of Pt atoms surrounded by Ti atoms at the center of the hexagons (Fig. 16c). For SA- and NC-based catalysts, coordination bonds are the predominant force in the metal–MXene interaction. Carbon atoms in MXenes act as ligands to provide bonding electrons, while supported metals provide empty orbitals to form stable covalent bond structures. For instance, on Cu/Mo₂CT_x/SiO₂, Cu atoms located on the edges of Mo₂CT_x were speculated to be connected with MXenes through coordination.¹²² It was likely that surface μ-oxo sites (Mo–O–Mo) of Mo₂CT_x interacted with Cu⁰ by forming Mo–O–Cu⁺ linkages (Fig. 16d). Similarly, in Pt₁/Ti_3−xC₂T_x, the single Pt atoms were stabilized in the Ti vacancies and then coordinated with three carbon atoms on the MXene nanosheets.¹³⁵

Fig. 16 The metal–support interactions between transition metals and MXenes. (a) Schematic illustration of the preparation of MXenes from MAX via HF/HCl etching and loading Pd₅₀–Ru₅₀ TMNPs by electrostatic interaction. Reproduced with permission.¹³² Copyright 2021, Elsevier. (b) The (111) surface of Pt₃Ti NP. (Inset) A simulated STEM image of the Pt₃Ti (111) surface. (c) Schematic of RMSI in Pt/MXene catalysts and the structure of L1₂-ordered intermetallic Pt₃Ti. Reproduced with permission.¹³⁶ Copyright 2018, Springer Nature. (d) The synthesis of Cu/Mo₂CT_x/SiO_2-500 with highly dispersed Cu sites in coordination interaction with partially reduced Mo₂CT_x nanosheets. Reproduced with permission.¹²² Copyright 2021, Springer Nature.

Strong metal–MXene interaction can promote the stability of the catalysts, and meanwhile, the electronic properties of MXene-supported catalysts are modified, for the reason that the interfaces of the stable metal–MXenes catalysts provide a bridge for electron transfer and rearrangement in both transition metals and MXene carriers, thus forming more active catalysts. For instance, the interaction between the Cu₃ cluster and monolayer defective Mo₂CO₂ (d-Mo₂CO₂) made Cu₃ clusters act as an electron reservoir to boost the catalysis of CO oxidation.⁴⁸ Before O₂ adsorption, the doped Cu₃ cluster functioned as an electron reservoir to collect electrons from defective Mo₂CO₂, while it donated electrons when adsorbing O₂ and CO. To be precise, some electrons were transferred from Cu atoms to occupy the empty 2π*-antibonding orbits of gas-phase O₂ and CO, resulting in lower adsorption energy of O₂ and CO. In addition, the variation of electronic properties also occurred on Pt/Ti₃C₂T_x when it acted as an efficient catalyst for light alkane dehydrogenation (LADH).¹⁴³ During the synthesis of Pt/Ti₃C₂T_x, Pt₃Ti intermetallic phase would be formed with a lower d band center. This change resulted in the weaker adsorption of light hydrocarbon species and changed the relative free energy and barriers of the reaction steps of LADH.

5. Summary and outlook

In summary, MXenes with a robust M_n+1X_n skeleton and tunable surface termination groups have gradually attracted interest from the researchers in thermal catalysis, due to their noble-metal-like catalytic activity and high thermal stability. Among them, Ti₃C₂T_x, Mo₂CT_x, and V₂CT_x are the three most widely studied MXenes in thermal catalysis, which can usually be obtained through a HF selective etching process. They can act as catalysts to directly participate in thermal reactions like hydrogenation, dehydrogenation, CO oxidation, etc. Their catalytic performance, such as stability and selectivity, can be adjusted by varying the types and the coverage of the surface termination groups, owing to the role of T_x groups in adjacent atoms and surface properties. Meanwhile, the efficiency of the reaction can also be directly boosted through the unique design of the active sites on the surface of MXenes, which would mainly facilitate the decrease of the energy barriers of the reactants. Besides, MXenes can also serve as catalytic carriers to support TMs in the form of NPs, NCs, and SAs, which can further improve the stability of the catalysts compared with traditional metal-based catalysts. In particularly, through metal–support interaction including electrostatic interaction, intermetallic interaction, and coordination interaction, a bridge between TMs and MXenes can be built to transfer the electrons, resulting in the rearrangement of the electrons in catalysts and thus improving the catalytic activity.¹⁵⁴

Although MXenes in thermal catalysis have been investigated extensively, these materials are still facing some obstacles in practical applications. One of the major problems is the long-term stability, because MXenes are quite sensitive to both water and oxygen. Meanwhile, considering that harsh conditions like high temperature and high pressure are usually required in thermal catalysis, the acceleration of the deactivation of MXenes will commonly occur. Therefore, the synthesis of extremely stable MXene-based catalysts, which can catalyze reactions under milder conditions, should be explored. Moreover, the cost and the security in the synthesis of MXenes are worthy of attention, and low cost and environmentally friendly fluorine-free preparation routes instead of the HF etching method for MXene synthesis are desired. More importantly, there is still a lack of standards and protocols that could point out how to design more efficient MXene catalysts. Currently, DFT is widely applied to design and assess the structure of catalysts before the actual catalysis is performed.³⁸ However, the DFT investigation of MXenes in thermal catalysis is rather limited. More modelling and calculations ought to be performed to systematically forecast the catalytic performance of MXenes, which can better guide the preparation of MXenes-based catalysts.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the Huaneng Group Science and Technology Research Project (KTHT-U22YYJC12), Tsinghua University-Shanxi Clean Energy Research Institute Innovation Project Seed Fund, and Tsinghua University Initiative Scientific Research Program (20213080042).

References

1. X. Jiang, A. V. Kuklin, A. Baev, Y. Ge, H. Ågren, H. Zhang and P. N. Prasad, Phys. Rep., 2020, 848, 1–58.
2. Y. Gogotsi and B. Anasori, ACS Nano, 2019, 13, 8491–8494.
3. Y. Wang, J. Mao, X. Meng, L. Yu, D. Deng and X. Bao, Chem. Rev., 2019, 119, 1806–1854.
4. Y. Chen, Z. Fan, Z. Zhang, W. Niu, C. Li, N. Yang, B. Chen and H. Zhang, Chem. Rev., 2018, 118, 6409–6455.
5. M. Liu, P. Zhang, Z. Qu, Y. Yan, C. Lai, T. Liu and S. Zhang, Nat. Commun., 2019, 10, 3917.
6. Y. Yan, P. Zhang, Z. Qu, M. Tong, S. Zhao, Z. Li, M. Liu and Z. Lin, Nano Lett., 2020, 20, 7662–7669.
7. Y. Yan, S. Liang, X. Wang, M. Zhang, S. Hao, X. Cui, Z. Li and Z. Lin, Proc. Natl. Acad. Sci. U. S. A., 2021, 118, e2110036118.
8. X. Cai, Y. Luo, B. Liu and H. M. Cheng, Chem. Soc. Rev., 2018, 47, 6224–6266.
9. Y. Xue, G. Zhao, R. Yang, F. Chu, J. Chen, L. Wang and X. Huang, Nanoscale, 2021, 13, 3911–3936.
10. S. Shirvani, M. Ghashghaee and K. J. Smith, Catal. Rev., 2021, 1–51,  DOI:10.1080/01614940.2021.1899605.
11. K. Hantanasirisakul and Y. Gogotsi, Adv. Mater., 2018, 30, 1804779.
12. M. Naguib, O. Mashtalir, J. Carle, V. Presser, J. Lu, L. Hultman, Y. Gogotsi and M. W. Barsoum, ACS Nano, 2012, 6, 1322–1331.
13. B. Anasori, M. R. Lukatskaya and Y. Gogotsi, Nat. Rev. Mater., 2017, 2, 16098.
14. J. Lei, X. Zhang and Z. Zhou, Front. Phys., 2015, 10, 276–286.
15. N. K. Chaudhari, H. Jin, B. Kim, D. San Baek, S. H. Joo and K. Lee, J. Mater. Chem. A, 2017, 5, 24564–24579.
16. I. M. Chirica, A. G. Mirea, T. Nea U, M. Florea, M. W. Barsoum and F. Nea U, J. Mater. Chem. A, 2021, 9, 19589–19612.
17. M. Shekhirev, C. E. Shuck, A. Sarycheva and Y. Gogotsi, Prog. Mater. Sci., 2021, 120, 100757.
18. J. Zhu, E. Ha, G. Zhao, Y. Zhou, D. Huang, G. Yue, L. Hu, N. Sun, Y. Wang, L. Y. S. Lee, C. Xu, K. Wong, D. Astruc and P. Zhao, Coord. Chem. Rev., 2017, 352, 306–327.
19. P. O. Å. Persson and J. Rosen, Curr. Opin. Solid State Mater. Sci., 2019, 23, 100774.
20. A. Agresti, A. Pazniak, S. Pescetelli, A. Di Vito, D. Rossi, A. Pecchia, M. Auf Der Maur, A. Liedl, R. Larciprete, D. V. Kuznetsov, D. Saranin and A. Di Carlo, Nat. Mater., 2019, 18, 1228–1234.
21. B. Ahmed, A. E. Ghazaly and J. Rosen, Adv. Funct. Mater., 2020, 30, 2000894.
22. F. Song, G. Li, Y. Zhu, Z. Wu, X. Xie and N. Zhang, J. Mater. Chem. A, 2020, 8, 18538–18559.
23. C. Feng, Z. P. Wu, K. W. Huang, J. Ye and H. Zhang, Adv. Mater., 2022, 2200180.
24. M. M. Hasan, M. M. Hossain and H. K. Chowdhury, J. Mater. Chem. A, 2021, 9, 3231–3269.
25. X. Zhan, C. Si, J. Zhou and Z. Sun, Nanoscale Horiz., 2020, 5, 235–258.
26. D. Xiong, Y. Shi and H. Y. Yang, Mater. Today, 2021, 46, 183–211.
27. D. Johnson, Z. Qiao, E. Uwadiunor and A. Djire, Small, 2022, 18, 2106129.
28. Q. Tang, Z. Zhou and P. Shen, J. Am. Chem. Soc., 2012, 134, 16909–16916.
29. X. Li, C. Wang, Y. Cao and G. Wang, Chem.–Asian J., 2018, 13, 2742–2757.
30. F. Shahzad, M. Alhabeb, C. B. Hatter, B. Anasori, H. S. Man, C. M. Koo and Y. Gogotsi, Science, 2016, 353, 1137–1140.
31. S. Kim, F. Gholamirad, M. Yu, C. M. Park, A. Jang, M. Jang, N. Taheri-Qazvini and Y. Yoon, Chem. Eng. J., 2021, 406, 126789.
32. X. Meng, X. Cui, N. P. Rajan, L. Yu, D. Deng and X. Bao, Chem, 2019, 5, 2296–2325.
33. N. H. Ahmad Junaidi, W. Y. Wong, K. S. Loh, S. Rahman and W. R. W. Daud, Int. J. Energy Res., 2021, 45, 15760–15782.
34. H. Wang, Y. Wu, X. Yuan, G. Zeng, J. Zhou, X. Wang and J. W. Chew, Adv. Mater., 2018, 30, 1704561.
35. Q. Zhong, Y. Li and G. Zhang, Chem. Eng. J., 2021, 409, 128099.
36. H. Huang, X. Jiang, N. Li, D. Chen, Q. Xu, H. Li, J. He and J. Lu, Appl. Catal., B, 2021, 284, 119754.
37. Q. Zhu, Y. Cui, Y. Zhang, Z. Cao, Y. Shi, J. Gu, Z. Du, B. Li and S. Yang, Mater. Today Nano, 2021, 13, 100104.
38. Á. Morales-García, F. Calle-Vallejo and F. Illas, ACS Catal., 2020, 10, 13487–13503.
39. R. B. Levy and M. Boudart, Science, 1973, 181, 547–549.
40. J. H. Sinfelt and D. J. C. Yates, Nat. Phys. Sci., 1971, 229, 27–28.
41. T. Hou, Q. Luo, Q. Li, H. Zu, P. Cui, S. Chen, Y. Lin, J. Chen, X. Zheng, W. Zhu, S. Liang, J. Yang and L. Wang, Nat. Commun., 2020, 11, 4251.
42. Z. Li and Y. Wu, Small, 2019, 15, 1804736.
43. J. D. Gouveia, Á. Morales-García, F. Viñes, J. R. B. Gomes and F. Illas, ACS Catal., 2020, 10, 5049–5056.
44. Q. Zhang, J. He, X. Fu, S. Xie, R. Fan, H. Lai, W. Cheng, P. Ji, J. Sheng, Q. Liao, W. Zhu and H. Li, Chem. Eng. J., 2022, 430, 132950.
45. D. A. Kuznetsov, Z. Chen, P. V. Kumar, A. Tsoukalou, A. Kierzkowska, P. M. Abdala, O. V. Safonova, A. Fedorov and C. R. Müller, J. Am. Chem. Soc., 2019, 141, 17809–17816.
46. X. Yang, Z. Lu, C. Cheng, Y. Wang, X. Zhang, Z. Yang and W. Lu, J. Phys. Chem. C, 2020, 124, 4090–4100.
47. K. Li, T. Jiao, R. Xing, G. Zou, J. Zhou, L. Zhang and Q. Peng, Sci. China Mater., 2018, 61, 728–736.
48. C. Cheng, X. Zhang, Z. Yang and Z. Zhou, ACS Appl. Mater. Interfaces, 2018, 10, 32903–32912.
49. S. H. Talib, S. Baskaran, X. Yu, Q. Yu, B. Bashir, S. Muhammad, S. Hussain, X. Chen and J. Li, Sci. China Mater., 2021, 64, 651–663.
50. S. Samanta and R. Srivastava, Mater. Adv., 2020, 1, 1506–1545.
51. H. Zhou, Z. Chen, E. Kountoupi, A. Tsoukalou, P. M. Abdala, P. Florian, A. Fedorov and C. R. Müller, Nat. Commun., 2021, 12, 5510.
52. L. H. Zhang, Y. Shi, Y. Wang and N. R. Shiju, Adv. Sci., 2020, 7, 1902126.
53. J. Nan, X. Guo, J. Xiao, X. Li, W. Chen, W. Wu, H. Liu, Y. Wang, M. Wu and G. Wang, Small, 2021, 17, 1902085.
54. J. Halim, J. Palisaitis, J. Lu, J. Thörnberg, E. J. Moon, M. Precner, P. Eklund, P. O. Å. Persson, M. W. Barsoum and J. Rosen, ACS Appl. Nano Mater., 2018, 1, 2455–2460.
55. L. Wang, M. Han, C. E. Shuck, X. Wang and Y. Gogotsi, Nano Energy, 2021, 88, 106308.
56. W. Hong, B. C. Wyatt, S. K. Nemani and B. Anasori, MRS Bull., 2020, 45, 850–861.
57. B. Ahmed, A. E. Ghazaly and J. Rosen, Adv. Funct. Mater., 2020, 30, 2000894.
58. J. L. Hart, K. Hantanasirisakul, A. C. Lang, B. Anasori, D. Pinto, Y. Pivak, J. T. van Omme, S. J. May, Y. Gogotsi and M. L. Taheri, Nat. Commun., 2019, 10, 522.
59. L. Gao, W. Bao, A. V. Kuklin, S. Mei, H. Zhang and H. Ågren, Adv. Mater., 2021, 33, 2004129.
60. M. Naguib, M. W. Barsoum and Y. Gogotsi, Adv. Mater., 2021, 33, 2103393.
61. M. Naguib, M. Kurtoglu, V. Presser, J. Lu, J. Niu, M. Heon, L. Hultman, Y. Gogotsi and M. W. Barsoum, Adv. Mater., 2011, 23, 4248–4253.
62. P. Lakhe, E. M. Prehn, T. Habib, J. L. Lutkenhaus, M. Radovic, M. S. Mannan and M. J. Green, Ind. Eng. Chem. Res., 2019, 58, 1570–1579.
63. M. Ghidiu, M. R. Lukatskaya, M. Zhao, Y. Gogotsi and M. W. Barsoum, Nature, 2014, 516, 78–81.
64. C. Peng, P. Wei, X. Chen, Y. Zhang, F. Zhu, Y. Cao, H. Wang, H. Yu and F. Peng, Ceram. Int., 2018, 44, 18886–18893.
65. L. Wang, H. Zhang, B. Wang, C. Shen, C. Zhang, Q. Hu, A. Zhou and B. Liu, Electron. Mater. Lett., 2016, 12, 702–710.
66. J. Halim, M. R. Lukatskaya, K. M. Cook, J. Lu, C. R. Smith, L. Näslund, S. J. May, L. Hultman, Y. Gogotsi, P. Eklund and M. W. Barsoum, Chem. Mater., 2014, 26, 2374–2381.
67. Y. Wei, P. Zhang, R. A. Soomro, Q. Zhu and B. Xu, Adv. Mater., 2021, 33, 2103148.
68. M. Naguib, J. Halim, J. Lu, K. M. Cook, L. Hultman, Y. Gogotsi and M. W. Barsoum, J. Am. Chem. Soc., 2013, 135, 15966–15969.
69. M. Naguib and Y. Gogotsi, Acc. Chem. Res., 2015, 48, 128–135.
70. P. Urbankowski, B. Anasori, T. Makaryan, D. Er, S. Kota, P. L. Walsh, M. Zhao, V. B. Shenoy, M. W. Barsoum and Y. Gogotsi, Nanoscale, 2016, 8, 11385–11391.
71. M. Li, J. Lu, K. Luo, Y. Li, K. Chang, K. Chen, J. Zhou, J. Rosen, L. Hultman, P. Eklund, P. O. Å. Persson, S. Du, Z. Chai, Z. Huang and Q. Huang, J. Am. Chem. Soc., 2019, 141, 4730–4737.
72. G. Li, L. Tan, Y. Zhang, B. Wu and L. Li, Langmuir, 2017, 33, 9000–9006.
73. S. Yang, P. Zhang, F. Wang, A. G. Ricciardulli, M. R. Lohe, P. W. M. Blom and X. Feng, Angew. Chem., Int. Ed., 2018, 57, 15491–15495.
74. J. Mei, G. A. Ayoko, C. Hu, J. M. Bell and Z. Sun, Sustainable Mater. Technol., 2020, 25, e156.
75. K. Huang, Z. Li, J. Lin, G. Han and P. Huang, Chem. Soc. Rev., 2018, 47, 5109–5124.
76. Y. Li, H. Shao, Z. Lin, J. Lu, L. Liu, B. Duployer, P. O. Å. Persson, P. Eklund, L. Hultman, M. Li, K. Chen, X. Zha, S. Du, P. Rozier, Z. Chai, E. Raymundo-Piñero, P. Taberna, P. Simon and Q. Huang, Nat. Mater., 2020, 19, 894–899.
77. V. Kamysbayev, A. S. Filatov, H. Hu, X. Rui, F. Lagunas, D. Wang, R. F. Klie and D. V. Talapin, Science, 2020, 369, 979–983.
78. T. Li, L. Yao, Q. Liu, J. Gu, R. Luo, J. Li, X. Yan, W. Wang, P. Liu, B. Chen, W. Zhang, W. Abbas, R. Naz and D. Zhang, Angew. Chem., Int. Ed., 2018, 57, 6115–6119.
79. G. Zou, Q. Zhang, C. Fernandez, G. Huang, J. Huang and Q. Peng, ACS Nano, 2017, 11, 12219–12229.
80. W. Sun, S. A. Shah, Y. Chen, Z. Tan, H. Gao, T. Habib, M. Radovic and M. J. Green, J. Mater. Chem. A, 2017, 5, 21663–21668.
81. M. R. Lukatskaya, J. Halim, B. Dyatkin, M. Naguib, Y. S. Buranova, M. W. Barsoum and Y. Gogotsi, Angew. Chem., Int. Ed., 2014, 53, 4877–4880.
82. X. Xie, Y. Xue, L. Li, S. Chen, Y. Nie, W. Ding and Z. Wei, Nanoscale, 2014, 6, 11035–11040.
83. S. Pang, Y. Wong, S. Yuan, Y. Liu, M. Tsang, Z. Yang, H. Huang, W. Wong and J. Hao, J. Am. Chem. Soc., 2019, 141, 9610–9616.
84. T. Yin, Y. Li, R. Wang, O. A. Al-Hartomy, A. Al-Ghamdi, S. Wageh, X. Luo, X. Tang and H. Zhang, Ceram. Int., 2021, 47, 28642–28649.
85. H. Shi, P. Zhang, Z. Liu, S. Park, M. R. Lohe, Y. Wu, A. Shaygan Nia, S. Yang and X. Feng, Angew. Chem., Int. Ed., 2021, 60, 8689–8693.
86. W. Meng, X. Liu, H. Song, Y. Xie, X. Shi, M. Dargusch, Z. Chen, Z. Tang and S. Lu, Nano Today, 2021, 40, 101273.
87. L. Verger, V. Natu, M. Carey and M. W. Barsoum, Trends Chem., 2019, 1, 656–669.
88. T. Yin, Y. Li, R. Wang, O. A. Al-Hartomy, A. Al-Ghamdi, S. Wageh, X. Luo, X. Tang and H. Zhang, Ceram. Int., 2021, 47, 28642–28649.
89. Q. Tao, M. Dahlqvist, J. Lu, S. Kota, R. Meshkian, J. Halim, J. Palisaitis, L. Hultman, M. W. Barsoum, P. O. Å. Persson and J. Rosen, Nat. Commun., 2017, 8, 14949.
90. M. Alhabeb, K. Maleski, T. S. Mathis, A. Sarycheva, C. B. Hatter, S. Uzun, A. Levitt and Y. Gogotsi, Angew. Chem., Int. Ed., 2018, 57, 5444–5448.
91. M. Alhabeb, K. Maleski, B. Anasori, P. Lelyukh, L. Clark, S. Sin and Y. Gogotsi, Chem. Mater., 2017, 29, 7633–7644.
92. B. Wang, M. Wang, F. Liu, Q. Zhang, S. Yao, X. Liu and F. Huang, Angew. Chem., Int. Ed., 2020, 59, 1914–1918.
93. M. Naguib, W. Tang, K. L. Browning, G. M. Veith, V. Maliekkal, M. Neurock and A. Villa, ChemCatChem, 2020, 12, 5733–5742.
94. E. Blanco, A. Rosenkranz, R. Espinoza-González, V. M. Fuenzalida, Z. Zhang, S. Suárez and N. Escalona, Catal. Commun., 2020, 133, 105833.
95. J. Diao, M. Hu, Z. Lian, Z. Li, H. Zhang, F. Huang, B. Li, X. Wang, D. S. Su and H. Liu, ACS Catal., 2018, 8, 10051–10057.
96. J. He, P. Wu, L. Chen, H. Li, M. Hua, L. Lu, Y. Wei, Y. Chao, S. Zhou, W. Zhu and H. Li, Chem. Eng. J., 2021, 416, 129022.
97. R. Thakur, A. VahidMohammadi, J. Smith, M. Hoffman, J. Moncada, M. Beidaghi and C. A. Carrero, ACS Catal., 2020, 10, 5124–5134.
98. A. Kurlov, E. B. Deeva, P. M. Abdala, D. Lebedev, A. Tsoukalou, A. Comas-Vives, A. Fedorov and C. R. Müller, Nat. Commun., 2020, 11, 4920.
99. R. Thakur, M. Hoffman, A. VahidMohammadi, J. Smith, M. Chi, B. Tatarchuk, M. Beidaghi and C. A. Carrero, ChemCatChem, 2020, 12, 3639–3643.
100. E. B. Deeva, A. Kurlov, P. M. Abdala, D. Lebedev, S. M. Kim, C. P. Gordon, A. Tsoukalou, A. Fedorov and C. R. Müller, Chem. Mater., 2019, 31, 4505–4513.
101. A. D. Dillon, M. J. Ghidiu, A. L. Krick, J. Griggs, S. J. May, Y. Gogotsi, M. W. Barsoum and A. T. Fafarman, Adv. Funct. Mater., 2016, 26, 4162–4168.
102. K. Maleski, V. N. Mochalin and Y. Gogotsi, Chem. Mater., 2017, 29, 1632–1640.
103. M. J. Gilkey and B. Xu, ACS Catal., 2016, 6, 1420–1436.
104. C. Wang, L. Wang, J. Zhang, H. Wang, J. P. Lewis and F. Xiao, J. Am. Chem. Soc., 2016, 138, 7880–7883.
105. E. Blanco, C. Sepulveda, K. Cruces, J. L. García-Fierro, I. T. Ghampson and N. Escalona, Catal. Today, 2020, 356, 376–383.
106. M. M. Bhasin, J. H. McCain, B. V. Vora, T. Imai and P. R. Pujadó, Appl. Catal., A, 2001, 221, 397–419.
107. B. V. Vora, Top. Catal., 2012, 55, 1297–1308.
108. M. C. Denney, V. Pons, T. J. Hebden, D. M. Heinekey and K. I. Goldberg, J. Am. Chem. Soc., 2006, 128, 12048–12049.
109. R. T. Yang, A. J. Hernández-Maldonado and F. H. Yang, Science, 2003, 301, 79–81.
110. A. Rajendran, T. Cui, H. Fan, Z. Yang, J. Feng and W. Li, J. Mater. Chem. A, 2020, 8, 2246–2285.
111. C. Chen and A. Bhan, ACS Catal., 2017, 7, 1113–1122.
112. M. D. Porosoff, X. Yang, J. A. Boscoboinik and J. G. Chen, Angew. Chem., Int. Ed., 2014, 53, 6705–6709.
113. J. B. Claridge, A. P. E. York, A. J. Brungs, C. Marquez-Alvarez, J. Sloan, S. C. Tsang and M. L. H. Green, J. Catal., 1998, 180, 85–100.
114. C. Shi, A. Zhang, X. Li, S. Zhang, A. Zhu, Y. Ma and C. Au, Appl. Catal., A, 2012, 431–432, 164–170.
115. F. Solymosi, R. Németh, L. Óvári and L. Egri, J. Catal., 2000, 195, 316–325.
116. C. Chen and A. Bhan, ACS Catal., 2017, 7, 1113–1122.
117. M. M. Sullivan and A. Bhan, ACS Catal., 2016, 6, 1145–1152.
118. H. Bian, D. Li, S. Wang, J. Yan and S. F. Liu, Chem. Sci., 2022, 13, 1335–1341.
119. J. Chen, Y. Xu, P. Liao, H. Wang and H. Zhou, Carbon Capture Sci. Technol., 2022, 4, 100052.
120. C. Shen, K. Sun, Z. Zhang, N. Rui, X. Jia, D. Mei and C. Liu, ACS Catal., 2021, 11, 4036–4046.
121. Y. Xu, C. Luo, H. Sang, B. Lu, F. Wu, X. Li and L. Zhang, Chem. Eng. J., 2022, 435, 134960.
122. H. Zhou, Z. Chen, A. V. López, E. D. López, E. Lam, A. Tsoukalou, E. Willinger, D. A. Kuznetsov, D. Mance, A. Kierzkowska, F. Donat, P. M. Abdala, A. Comas-Vives, C. Copéret, A. Fedorov and C. R. Müller, Nat. Catal., 2021, 4, 860–871.
123. D. S. Newsome, Catal. Rev., 1980, 21, 275–318.
124. C. Ratnasamy and J. P. Wagner, Catal. Rev., 2009, 51, 325–440.
125. J. A. Rodriguez, S. Ma, P. Liu, J. Hrbek, J. Evans and M. Pérez, Science, 2007, 318, 1757–1760.
126. R. Thakur, M. Hoffman, A. VahidMohammadi, J. Smith, M. Chi, B. Tatarchuk, M. Beidaghi and C. A. Carrero, Chemcatchem, 2020, 12, 3639–3643.
127. T. Zambelli, J. Wintterlin, J. Trost and G. Ertl, Science, 1996, 273, 1688–1690.
128. K. Han, X. Zhang, P. Deng, Q. Jiao and E. Chu, Vacuum, 2020, 180, 109572.
129. J. D. Gouveia, Á. Morales-García, F. Viñes, F. Illas and J. R. B. Gomes, Appl. Catal., B, 2020, 260, 118191.
130. Y. Zhang, X. Zhang, C. Cheng and Z. Yang, Chin. Chem. Lett., 2020, 31, 931–936.
131. Y. Wang, J. Wang, G. Han, C. Du, Q. Deng, Y. Gao, G. Yin and Y. Song, Ceram. Int., 2019, 45, 2411–2417.
132. G. Bharath, K. Rambabu, A. Hai, I. Othman, N. Ponpandian, F. Banat and P. Loke Show, Chem. Eng. J., 2021, 414, 128869.
133. M. Sankar, Q. He, R. V. Engel, M. A. Sainna, A. J. Logsdail, A. Roldan, D. J. Willock, N. Agarwal, C. J. Kiely and G. J. Hutchings, Chem. Rev., 2020, 120, 3890–3938.
134. J. Ma, Q. Jiang, Y. Zhou, W. Chu, S. Perathoner, C. Jiang, K. H. Wu, G. Centi and Y. Liu, Small, 2021, 17, 2007509.
135. D. Zhao, Z. Chen, W. Yang, S. Liu, X. Zhang, Y. Yu, W. Cheong, L. Zheng, F. Ren, G. Ying, X. Cao, D. Wang, Q. Peng, G. Wang and C. Chen, J. Am. Chem. Soc., 2019, 141, 4086–4093.
136. Z. Li, L. Yu, C. Milligan, T. Ma, L. Zhou, Y. Cui, Z. Qi, N. Libretto, B. Xu, J. Luo, E. Shi, Z. Wu, H. Xin, W. N. Delgass, J. T. Miller and Y. Wu, Nat. Commun., 2018, 9, 5258.
137. Z. Li, Y. Cui, Z. Wu, C. Milligan, L. Zhou, G. Mitchell, B. Xu, E. Shi, J. T. Miller, F. H. Ribeiro and Y. Wu, Nat. Catal., 2018, 1, 349–355.
138. Z. Li, Y. Xiao, P. R. Chowdhury, Z. Wu, T. Ma, J. Z. Chen, G. Wan, T. Kim, D. Jing, P. He, P. J. Potdar, L. Zhou, Z. Zeng, X. Ruan, J. T. Miller, J. P. Greeley, Y. Wu and A. Varma, Nat. Catal., 2021, 4, 882–891.
139. L. D. Pachón and G. Rothenberg, Appl. Organomet. Chem., 2008, 22, 288–299.
140. N. Yan, C. Xiao and Y. Kou, Coord. Chem. Rev., 2010, 254, 1179–1218.
141. M. Liu, Z. Lyu, Y. Zhang, R. Chen, M. Xie and Y. Xia, Nano Lett., 2021, 21, 2248–2254.
142. W. A. Spieker and J. R. Regalbuto, Chem. Eng. Sci., 2001, 56, 3491–3504.
143. Q. Chen, W. Jiang and G. Fan, Dalton Trans., 2020, 49, 14914–14920.
144. T. K. Slot, F. Yue, H. Xu, E. V. Ramos-Fernandez, A. Sepúlveda-Escribano, Z. Sofer, G. Rothenberg and N. R. Shiju, 2D Mater., 2020, 8, 15001.
145. T. Liu, Z. Zhang, L. Yan, Z. Zhang, Y. Zhang and Y. Yin, Int. J. Hydrogen Energy, 2021, 46, 33098–33106.
146. J. Liu, H. Xiao and J. Li, J. Am. Chem. Soc., 2020, 142, 3375–3383.
147. V. Ramalingam, P. Varadhan, H. C. Fu, H. Kim, D. Zhang, S. Chen, L. Song, D. Ma, Y. Wang, H. N. Alshareef and J. H. He, Adv. Mater., 2019, 31, 1903841.
148. Y. Yan, H. Cheng, Z. Qu, R. Yu, F. Liu, Q. Ma, S. Zhao, H. Hu, Y. Cheng, C. Yang, Z. Li, X. Wang, S. Hao, Y. Chen and M. Liu, J. Mater. Chem. A, 2021, 9, 19489–19507.
149. X. Wu, H. Zhang, S. Zuo, J. Dong, Y. Li, J. Zhang and Y. Han, Nano-Micro Lett., 2021, 13, 1–28.
150. N. Cheng, S. Stambula, D. Wang, M. N. Banis, J. Liu, A. Riese, B. Xiao, R. Li, T. Sham, L. Liu, G. A. Botton and X. Sun, Nat. Commun., 2016, 7, 13638.
151. X. Zhang, J. Lei, D. Wu, X. Zhao, Y. Jing and Z. Zhou, J. Mater. Chem. A, 2016, 4, 4871–4876.
152. C. Cheng, X. Zhang, M. Wang, S. Wang and Z. Yang, Phys. Chem. Chem. Phys., 2018, 20, 3504–3513.
153. B. L. Mojet, J. T. Miller, D. E. Ramaker and D. C. Koningsberger, J. Catal., 1999, 186, 373–386.
154. S. Zhao, Y. Kang, M. Liu, B. Wen, Q. Fang, Y. Tang, S. He, X. Ma, M. Liu and Y. Yan, J. Mater. Chem. A, 2021, 9, 18927–18946.

---