Engineering Two-Dimensional Materials and Their Heterostructures as High-Performance Electrocatalysts
- 773 Downloads
Electrochemical energy conversion between electricity and chemicals through electrocatalysis is a promising strategy for the development of clean and sustainable energy sources. This is because efficient electrocatalysts can greatly reduce energy loss during the conversion process. However, poor catalytic performances and a shortage in catalyst material resources have greatly restricted the widespread applications of electrocatalysts in these energy conversion processes. To address this issue, earth-abundant two-dimensional (2D) materials with large specific surface areas and easily tunable electronic structures have emerged in recent years as promising high-performance electrocatalysts in various reactions, and because of this, this review will comprehensively discuss the engineering of these novel 2D material-based electrocatalysts and their associated heterostructures. In this review, the fundamental principles of electrocatalysis and important electrocatalytic reactions are introduced. Following this, the unique advantages of 2D material-based electrocatalysts are discussed and catalytic performance enhancement strategies are presented, including the tuning of electronic structures through various methods such as heteroatom doping, defect engineering, strain engineering, phase conversion and ion intercalation, as well as the construction of heterostructures based on 2D materials to capitalize on individual advantages. Finally, key challenges and opportunities for the future development of these electrocatalysts in practical energy conversion applications are presented.
Keywords2D materials Electrocatalysts Heterostructures Hydrogen evolution reaction Oxygen reduction reaction Oxygen evolution reaction
The start of the second industrial revolution in the nineteenth century led to the invention and widespread use of the steam engine and as a result produced unprecedented demands for energy. Initially, this demand was met by thermal energy (heat) generated from wood or coal, but these resources suffered from low energy conversion and harvesting efficiencies. Later, fossil fuels were used to produce mechanical energy in the pursuit of higher energy conversion efficiencies. But efficiencies remain unsatisfactory. In addition, the excess use of fossil fuels has led to fossil fuel shortages and serious environmental problems. As a result, the development of sustainable and renewable energy resources is urgently needed [1, 2]. And because traditional energy conversion methods are unsatisfactory due to their low conversion efficiencies, the development of efficient, clean energy storage and conversion technologies is also crucial. Based on this, great efforts have been made in recent years to develop sustainable energy sources such as solar and hydrogen energy. Moreover, numerous efforts have been made in the efficient conversion of these sustainable energy sources to electrical energy [3, 4, 5] via energy conversion devices such as water electrolysis, fuel cells, and metal-air batteries [5, 6, 7]. And because these efficient energy conversion processes require electrocatalytic reactions, electrocatalysis is the key in the successful application of these energy conversion technologies . One obvious advantage of electrocatalysis is that electrocatalysts can lower reaction energy barriers to allow chemical reactions to occur under smaller electric fields (overpotential), leading to lower energy consumption during the conversion process. In addition, the use of electrocatalysts can allow for the easy control of reaction pathways and rates to produce products.
Overall, electrochemical reactions occur at the interface between electrocatalysts and reactants in solution in which the surface of electrocatalysts interacts with dissolved reactants. Here, the performance of electrocatalysts is mainly determined by factors such as chemical composition, crystal structure, microstructure, and macromorphology, and these parameters have a direct influence on catalytic performance, meaning that catalytically active interfaces are vital for high-efficiency electrocatalysis . A key consideration of interfacial catalysis is the adsorption and desorption ability of reaction intermediates by the catalyst in which desirable catalysts must possess suitable adsorption and desorption energy for intermediates to facilitate electrocatalytic reactions. Other considerations include good electron conductivity, rapid mass transfer, high durability, and environmental benignity [10, 11, 12]. Because of this, not only the intrinsic activity of catalysts is important, but the structure of catalysts is also critical to achieve efficient catalytic processes in which a major goal of electrocatalysis research is the design and synthesis of materials with high activity, high selectivity, and good durability. Therefore, investigations into the relationship between catalyst structure and catalytic performance are important because interfacial structures have direct influences on efficiency and selectivity.
Currently, noble metals (e.g., Pt) and noble metal oxides (e.g., RuO2 and IrO2) are being widely used as electrocatalysts due to their high performances in electrochemical reactions; however, limited resources restrict widespread application. To address this, increasing attention has been paid in recent years to the exploration of abundant low-dimensional materials, especially 2D materials, for electrocatalysis due to their tunable electronic structures, large specific surface areas, and scalable production capabilities [13, 14]. All of them are desirable properties for promising large-scale electrocatalyst applications. Therefore, this review will mainly focus on the rational engineering of 2D materials and corresponding heterostructures in the production of highly efficient electrocatalysts. In addition, current enhancement strategies for the catalytic performance of 2D materials, including heteroatom doping, defect engineering, strain engineering, phase conversion, and ion intercalation, are discussed. Furthermore, the fabrication of heterostructures based on 2D materials and their advantages in electrocatalysis are also discussed, and the challenges of future 2D material-based electrocatalyst development are presented.
2 Key Electrocatalysis Processes and the Potential of 2D Materials
HER is the cathode reaction in water electrolysis and can be divided into two steps in acidic media  in which the first step is the Volmer step involving electrochemical hydrogen adsorption and the second step can be further divided into two types including the Heyrovsky step and the Tafel step . Here, the rate of HER depends on the hydrogen adsorption free energy (∆GH*)  in which if hydrogen adsorption on the catalytic surface is weak, hydrogen adsorption becomes the rate-determining step of the overall reaction. Alternatively, if hydrogen adsorption is too strong, the desorption step becomes the rate-determining step of the overall reaction. Based on this, efficient HER catalysts should possess a ∆GH* close to 0 eV, allowing for both the adsorption of reactants and the desorption of products to occur easily [18, 19, 20]. Furthermore, the Volmer step varies depending on the electrolyte used (acidic or alkaline) in which in acidic media, electrocatalysts can directly adsorb H+ from the solution, whereas in alkaline media, electrocatalysts need to dissociate H2O before being able to adsorb H+, which is more difficult to realize than the subsequent H+ reduction in HER for many electrocatalysts. Overall, there are several key parameters that can be used to evaluate the performance of HER electrocatalysts. These parameters include: (i) the electrochemical active surface area (ECSA), which is defined as the number of catalytically active sites on the catalyst; (ii) the onset potential, which is defined as the potential at which cathodic currents are observed; (iii) the overpotential (η), which is the additional potential required to drive a reaction to reach a certain current density (j); (iv) the Tafel slope, which is the slope between η and log |j| and can be used to analyze the rate-determining step; (v) the exchange current density (j0), which is a key descriptor to evaluate the catalytic efficiency; (vi) the turnover frequency (TOF), which is defined as the numbers of reactant molecules transformed per active site per second; (vii) the Faradic efficiency, which describes the efficiency of charge transfer in an electrochemical reaction system; and (viii) stability, which is a key parameter to evaluate long-term catalyst performance .
The ORR is also a cathode reaction that is important in fuel cells and metal-air batteries. ORR possesses two types of reaction mechanisms in which one is a dissociative mechanism and the other is an associative mechanism [19, 22]. Here, reaction mechanisms depend on the oxygen dissociation barrier of catalyst surfaces and the binding energy of oxygenated species on catalyst surfaces determines catalytic activity. Based on this, optimal ORR catalysts should possess moderate binding to intermediates (i.e., O* and OOH*) in which weak binding can limit electron or proton transfer to adsorbed oxygen and strong binding can lead to difficulties in H2O desorption, which in turn can block active site for further oxygen adsorption. Here, the underlying mechanism behind the different binding energies of intermediates and catalyst surfaces is related to their electronic structures , and catalysts with optimal binding energies can provide enhanced catalytic properties for ORR, similar to HER. In addition, other factors such as mass transfer and gas diffusion at the catalyst interface also influence the catalytic properties for ORR. Based on this, the engineering and optimization of the geometric structure of catalysts are effective strategies to develop desirable ORR electrocatalysts . As for the evaluation of ORR catalysts, the onset potential (Eonset) and the half-wave potential (E1/2, which is the potential corresponding to half of the limited current density) are generally used to provide quantitative evaluations of the ORR activity of electrocatalysts in which more positive Eonset and E1/2 values represent better catalytic performances. Other criteria used in the evaluation of ORR activity in catalysts include the kinetic current density (Jk), which can be obtained from the Levich–Koutecky equation , the electron transfer numbers and the amount of intermediate H2O2 produced, all of which can be obtained from rotating disk electrode (RDE) and rotating ring-disk electrode (RRDE) characterizations.
The OER is the anode reaction in water electrolysis and metal-air batteries and is the inverse of the ORR. OER also includes two reaction mechanisms based on the pH of the electrolyte [19, 26]. In acidic media, the first step of OER is the formation of water intermediates in which water molecules are adsorbed on the surface of the catalyst. These water intermediates subsequently release two electrons and two protons to form an O* intermediate. Following this, two different pathways can occur for the transformation of the O* intermediate to O2 in which one involves the direct combination of two O* to produce O2, whereas the other involves the formation of an OOH* intermediate through the combination of O* and H2O, leading to the formation of O2. Alternatively, the first step of OER in the alkaline media is the formation of OH− intermediates with the subsequent steps being similar to those in acidic media. And based on the analysis of the OER process, appropriate catalytic interfaces for the association and dissociation of intermediates are prerequisites for optimal OER electrocatalysts. A summary of the key reaction steps in HER, ORR, and OER is listed in Fig. 1.
In the initial phases of electrocatalysis research, noble metals were the most widely studied with Pt being one of the most active electrocatalysts for HER and ORR [27, 28] based on the fact that Pt-based catalysts possess low overpotentials for both HER (\(\leqslant\) 50 mV to achieve a current density of 10 mA cm−2) and ORR (~ 300 mV with limited current density) . However, unlike HER and ORR, OER requires large positive potentials; because metals are susceptible to oxidation at higher potentials, they are unsuitable. Alternatively, metal oxides and metal hydroxides generally possess decent OER performances [30, 31] and the use of noble metals and corresponding oxides for high-performance electrocatalysis is an active research area with significant progress being made in recent years [32, 33]. For example, despite large overpotentials (~ 300 mV to achieve a current density of 10 mA cm−2), iridium oxide (IrO2) and ruthenium oxide (RuO2) have become dominant OER catalysts in recent years . However, noble metals suffer from the disadvantages such as high costs and limited availability, which restrict their large-scale commercialization. Therefore, the development of high-performance catalysts composed of earth-abundant and cheap elements is needed to achieve breakthroughs in the field of electrochemical energy conversion. And in recent years, scientists have investigated electrocatalysts based on noble metals, non-noble metals and even novel metal-free electrocatalysts, including metal oxides (CoxOy, FexOy, MnxOy, NixOy) [24, 26, 35, 36], metal hydroxides (NiFe LDH, FeCo LDH) [37, 38], metal phosphides or nitrides (CoxPy, NixPy, MoxPy, MoxNy) [39, 40, 41, 42], and carbon-based materials (carbon nanotubes, graphene, graphdiyne) [43, 44, 45]. Here, researchers have demonstrated that non-noble metal and metal-free electrocatalysts can provide catalytic activities for different catalytic reactions; however, most possess poorer catalytic performances than noble metal-based catalysts. Therefore, alternative noble metal-free catalysts with high catalytic performances need to be developed for large-scale implementation.
Overall, many materials have been demonstrated to be active for electrocatalysis, and the design of ideal electrocatalysts with high activities and long stability includes many factors. First, ideal electrocatalysts should possess large specific surface areas with abundant exposed pores to maximize active sites. Second, the active sites of ideal electrocatalysts should possess suitable binding strengths to both reactants and products to optimize corresponding electrocatalytic reactions in which binding strengths can be tuned through the modulation of catalyst electronic structures. Third, ideal electrocatalysts should be able to conduct electrons to external circuits and accelerate electrode kinetics. Fourth, ideal electrocatalysts should be porous to promote mass transfer and gas diffusion in electrolytes. Lastly, ideal electrocatalytic materials should be earth-abundant and inexpensive as compared with noble metal catalysts.
Since the successful isolation of single-atom-thick graphene from graphite by Geim et al.  in 2004, 2D materials have attracted tremendous attention as electrocatalysts. Aside from graphene, there is also a large family of 2D materials including hexagonal boron nitride (h-BN), transition metal dichalcogenides (TMDCs), black phosphorus (BP), MXenes, layered double hydroxides (LDHs) and many others [47, 48, 49, 50] that have shown outstanding properties [51, 52, 53], ranging from insulators, semiconductors, metals to superconductors . 2D materials possess strong in-plane covalent bonding and weak interactions such as van der Waals forces between layers, allowing these materials to be applicable in electrochemical energy storage and conversion applications. In addition, 2D materials with varying exposed edge sites can be obtained by tuning lateral sizes, allowing for the tuning of catalytic performances [24, 55]. Other advantages of 2D materials include their lateral size and thickness as well as their extremely large specific surface areas. These features impart obvious advantages in electrocatalytic reactions in which 2D materials with large specific surface areas can form more active sites and atomic thicknesses can facilitate electron transfer. Furthermore, the electronic and interfacial structure of 2D materials can be tuned through different physical or chemical methods, which can provide active sites with proper adsorption/desorption abilities for intermediates in electrocatalytic reactions. Moreover, porous structures can be created through the enlargement of interlayer spacing in 2D materials, allowing for the formation of surface pores and the construction of heterostructures, which can accelerate mass transfer and gas diffusion, leading to increased catalytic performances. Lastly, as previously mentioned, there is a large family of 2D materials that contain many earth-abundant elements such as graphene and MoS2, making them promising for large-scale electrocatalysis applications. 2D materials are also ideal model electrocatalysts in the study of electrocatalytic processes because they possess clear and well-defined structures that are different from and simpler than nanoparticle catalysts, allowing for better investigations of electrocatalysis mechanisms. Furthermore, significant time and costs can be saved in computational studies for experimental investigations because 2D materials are easier to simulate as electrocatalysts for electrochemical reactions.
3 Rational Design of 2D Material-Based High-Performance Electrocatalysts
3.1 Individual 2D Materials for Electrocatalysis
3.1.1 Graphene and Their Derivates
Analogous to defects, heteroatom doping can also alter the inert electronic structure of graphene and heteroatom-doped graphene has attracted major interest as a promising substitute for noble metals in various electrocatalytic reactions. Dai et al.  were the first to report the heteroatom doping of carbon materials for electrocatalytic reactions in 2009, and Qu et al.  conducted a ground-breaking study of 2D materials with heteroatom doping through the synthesis of nitrogen-doped graphene (NG) using chemical vapor deposition (CVD) in 2010 in which the resulting NG as a cathode provided significantly enhanced electrocatalytic activities, long-term operational stability, and high tolerances to the crossover effect as compared with commercial Pt/C catalysts for ORR. Here, theoretical calculations indicate that the incorporation of nitrogen with higher electronegativity can generate substantial positive charge densities on adjacent carbon atoms, which can enhance the chemisorption of O2 at the interface with lower ORR overpotentials. Following this study, similar results for ORR were obtained for graphene doped with heteroatoms such as boron (B) , sulfur (S) , and phosphorus (P) . Here, sulfur- and phosphorus-doped graphene possesses similar effects to nitrogen-doped graphene because sulfur, phosphorus, and nitrogen all possess higher electronegativity than carbon. As for boron, although it possesses lower electronegativity than carbon, a similar principle can be used to explain the increased activity for boron-doped graphene as compared with pure graphene in which the carbon–boron bonds in boron-doped graphene can undergo slight polarizations to form local positive charges on boron sites, allowing boron sites to become the preferred sites for O2 adsorption for subsequent reactions . Furthermore, researchers have also developed dual-doped  and tri-doped  graphene as electrocatalysts with even better ORR performances due to synergistic effects. For example, Liang et al.  prepared N and S dual-doped graphene as an ORR catalyst and reported excellent catalytic activities with a high positive onset potential of −0.06 V and a large kinetic limiting current of 52 mA cm−2 at − 0.8 V (V vs. Ag/AgCl). Here, DFT calculations revealed that the synergistic performance improvements can be attributed to the redistribution of spin and charge densities caused by the dual doping of S and N atoms, which generated large numbers of C active sites. In another study, Zhang et al.  prepared a N, P and fluorine (F) tri-doped graphene as a multifunctional electrocatalyst and reported satisfactory electrocatalytic performances for ORR, OER, HER, and in a Zn-air battery.
As graphene derivates, 2D graphdiyne and g-C3N4 materials have also shown promising catalytic performances for electrocatalysis [74, 75]. Graphdiyne, possessing sp- and sp2-hybridized carbon atoms, a nano-porous structure , and good charge transport properties, has attracted development as high-performance electrocatalysts, and as a result, graphdiyne as metal-free electrocatalysts has shown excellent performances for both HER and ORR [77, 78, 79]. As another graphene derivates, g-C3N4, possessing a similar structure to graphene but with the higher N content, has also shown promising application in high-performance electrocatalysis. However, the conductivity of g-C3N4 is inferior to graphene, which hinders application. Therefore, scientists have made great efforts to improve the conductivity of g-C3N4 through combination with conductive materials such as graphene and carbon nanotubes [80, 81, 82, 83].
3.1.2 Transition Metal Dichalcogenides (TMDCs)
Although defects or dopants in graphene have shown great promise for electrocatalysis, they are generally difficult to control. Unlike graphene, however, other 2D materials such as TMDCs have been shown to be catalytically active toward HER even without modification despite weak activities [3, 31]. And in terms of catalytic properties, TMDCs possess many advantages as compared with graphene. First, TMDCs possess intrinsic activities for electrocatalytic reactions, whereas graphene is inert. Second, TMDCs possess different phases that are semiconducting or metallic and that can display different performances. Third, both transition metal atoms and nonmetal atoms can be substitutionally doped into TMDC lattices to provide more doping options as compared with graphene. And as a result of these advantages, significant efforts have been focused on TMDC catalysts in recent years [84, 85, 86].
Semiconducting TMDCs such as 2H-MoS2 have been shown to be active toward HER even though their bulk counterparts are not very active. Electrochemical hydrogen production by using MoS2 can be traced back to the 1970s . And after several decades of technological developments, MoS2 has become a potential substitute for commercial Pt/C catalysts in HER. The identification of active sites is essential to properly study catalytic activity. Based on this, Jaramillo et al.  systematically explored the surface sites of MoS2 nanoparticles both experimentally and theoretically in which MoS2 samples were synthesized by using physical vapor deposition, and the ratio of basal-plane sites to edge sites was controlled through sintering without changing the nature of the edges. Here, electrochemical tests revealed that the rate of the reaction was directly proportional to the number of edge sites for all samples rather than to the particle size of samples. To gain further insights into the catalytic nature of MoS2, the researchers in this study also conducted DFT calculations to identify the activity differences between basal planes and edge sites. Here, the ∆GH* of the edge sites was calculated to be + 0.08 eV, which was closer to 0 eV than that of the basal plan sites. These results also showed that the electrocatalytic activity of hydrogen evolution possessed a linear relationship with the number of edge sites of the MoS2 catalyst. And based on these results, a series of methods was developed to obtain large numbers of edge sites on MoS2 in which the most effective method is to create defects [89, 90, 91]. For example, Xie et al.  developed a scalable method to produce large numbers of defects in MoS2 surfaces to expose active edge sites and reported that the resulting MoS2 nanosheets possessed excellent HER activities with a small onset overpotential of 120 mV, a large cathodic current density, and a small Tafel slope of 50 mV dec−1, which is one of the best comprehensive performances reported for individual MoS2 electrocatalysts.
Enormous efforts have been made to expose more active sites in TMDCs to increase catalytic performance; however, these efforts are usually accompanied by losses of stability and electrical conductivity. In one study, Voiry et al.  reported that strain-induced local lattice distortions of WS2 resulted in the phase conversion of WS2 from a 2H to 1T enriched phase, which led to enhanced catalytic activities. The researchers in this study also performed first-principles calculations to reveal the mechanisms responsible for the high activity of 1T WS2 in which calculations suggested that strain can significantly increase the density of states near the Fermi energy and facilitate the binding of hydrogen to the surface of distorted 1T WS2. To further understand the role of strain, Tan et al.  prepared monolayer MoS2 on the internal surface of curved 3D nano-porous gold (NPG) in which side-view STEM images clearly indicated significant lattice distortions (Fig. 5d–g) and different S-Mo-S bonding angles (α) (Fig. 5h). Here, the researchers reported that the distorted MoS2 film with the out-of-plane strain demonstrated good electrocatalytic performances toward HER and that the activity of the MoS2 film decreased as sample thicknesses increased from monolayer to three layers (Fig. 5i) with the corresponding Tafel slopes increasing from 46 to 55 mV dec−1 (Fig. 5j). The researchers suggested that this difference in performance can be attributed to the change of S-Mo-S angles (α) in which if α is larger than 140°, the band gap of monolayer 2H MoS2 is close to 0 eV, indicating that local semiconductor-to-metal transition was caused by out-of-plane lattice bending. Furthermore, the calculations conducted in this study showed that the free energy is also related to α and that lattice bending can cause smaller binding free energy, explaining the better performance of curved monolayer MoS2. A similar result was obtained by Putungan et al.  who used biaxial tensile strain to study the effects of distortion in 1T phase group-VI TMDCs on the free energy of hydrogen adsorption. Here, the researchers found that the density of states near the Fermi level increases as strain is introduced, which causes a decrease in hydrogen adsorption energy. Furthermore, Li et al.  systematically investigated the effects of S-vacancies combined with strain on HER performance and concluded that this combination can lead to even higher HER activities. In addition, the researchers in this study also suggested that strain can effectively stabilize adsorbed hydrogen on the surface of basal planes. Overall, these experiments and theoretical calculations demonstrate that strain engineering can enhance the activity of TMDC catalysts.
Both strain and ion intercalation results show that the activities of metallic TMDCs are higher than those of semiconducting ones. Moreover, metallic TMDCs possess better conductivity than semiconducting TMDCs, allowing for faster electrode kinetics and electron transport, demonstrating their potential as electrocatalysts. An effective method to obtain metallic TMDC-based electrocatalysts is through direct synthesis using simple methods. For example, Yin et al.  synthesized partially crystallized 1T-MoSe2 nanosheets using a modified one-pot hydrothermal strategy in which 1T-MoSe2 nanosheets can be obtained as abundant NaBH4 reductants are consumed during the hydrothermal process. Here, the researchers reported that by increasing the reductant, the conversion of MoSe2 nanosheets from the 2H phase to the 1T phase can be enhanced, leading to improved HER activities. And as a result, the optimized MoSe2 possessed an η of 152 mV at a j of −10 mA cm−2. Despite these performances, however, this method cannot obtain fully crystallized metallic nanosheets, leading to low catalyst utilization. To address this, Liu et al.  recently conducted a significant study to investigate the active sites of 2D metallic group-5 MX2 (H-TaS2 and H-NbS2) electrocatalysts, which possess high basal-plane activities for HER as demonstrated through theoretical calculations and experiments. In this study, first-principles calculations were used to reveal the underlying electronic factors controlling basal-plane activity in which the researchers reported that the populated state of electrons near the lowest unoccupied state (εLUS) was the key parameter determining the hydrogen adsorption strength of MX2 surfaces and that εLUS values between − 6.4 eV and − 5.5 eV were usual, corresponding to − 0.5 eV H* < Ea < + 0.5 eV H*. Furthermore, the researchers suggested that among viable candidates, group-5 metal disulfides (H-VS2, H-NbS2 and H-TaS2) were the most promising for catalysts due to suitable εLUS (<− 5.8 eV) (Fig. 6f). In experiments, both H-NbS2 and H-TaS2 demonstrated predictably high performances toward HER with overpotentials of 50–60 mV at a current density of 10 mA cm−2 (Fig. 6h) and Tafel slopes close to 30 mV dec−1 (Fig. 6i), surpassing the performance of all other TMDCs. The researchers in this study also found that these materials exhibited unusual self-optimizing properties as they catalyzed hydrogen evolution. Here, the researchers suggested that this unusual self-optimization was caused by beneficial morphological changes that improved charge transfer and the accessibility of active sites in which H-NbS2 and H-TaS2 were both weakly bounded layered materials and hydrogen gas produced at the sites between the layers can cause exfoliation and fracture, resulting in the thinning and exposure of more active sites during HER (Fig. 6g). At the same time, Shi et al.  also reported that metallic 2H-TaS2 grown on gold foil was an efficient HER electrocatalyst in which the catalyst displayed thickness-dependent electrocatalytic performances. Here, a ~ 150-nm-thick 2H-TaS2 catalyst provided a Tafel slope of ~ 33 mV dec−1 that was similar to that of Pt (~ 31 mV dec−1), indicating a Volmer–Tafel mechanism for HER. However, although the catalytic activity of metallic TaS2 was superior to those of other 2D materials, its activity in the basal plane needs to be further improved to satisfy practical applications. Furthermore, metallic TMDCs are sensitive to air and water, and therefore, special attention needs to be paid to their stability. And overall, having only been applied in electrocatalysis recently, metallic TMDCs possessing abundant active sites on both edges and basal planes need to be developed further to gain optimized electrocatalytic performances.
3.1.3 Other 2D Materials
Although the exfoliation of LDHs can create more active sites to improve electrocatalytic activity, the intrinsic activity of each site remains unchanged and the improvement of individual site activity is of great interest. In one study, Liu et al.  reported a water-plasma-enabled exfoliation process that increased the number of active sites and improved intrinsic activity involving the treatment of pristine CoFe LDHs with water plasma to obtain ultrathin CoFe LDH nanosheets (Fig. 7e) in which TEM revealed the formation of a relatively rough surface with small irregular pores (Fig. 7f) and atomic-size vacancies after the plasma exfoliation process (Fig. 7g). And as a result, the ultrathin CoFe LDH nanosheet catalyst with a large number of vacancies exhibited excellent OER performances with a lower overpotential of 290 mV and a smaller Tafel slope of 36 mV dec−1 as compared with pristine CoFe LDHs (331 mV and 52 mV dec−1) at a current density of 10 mA cm−2 (Fig. 7h–i). Here, the researchers suggested that the vacancies in the ultrathin CoFe LDH nanosheets were responsible for the higher electrocatalytic activity as compared with perfect LDH nanosheets with no vacancies. In another study, Yang et al.  synthesized ultrathin CoFe LDH nanosheets with large numbers of defects as a bifunctional electrocatalyst for overall water splitting and reported high electrocatalytic activities for both OER and HER in alkaline media, demonstrating that the density of catalytic active sites in LDHs can be increased through defect engineering. A significant problem with LDHs is the limited number of active sites on their edges. In addition, the electrocatalytic mechanisms of LDHs remain unclear despite extensive efforts. And because of the limited number of active sites and the poor intrinsic activity of LDH-based electrocatalysts, widespread application is hindered in which novel and effective strategies such as cation doping or layer spacing modification need to be developed to resolve these issues.
Another interesting family of 2D materials are MXenes, which possess the general formula of Mn+1AXn in which M is an early transition metal (e.g., Mo and Ti), A is C and/or N, and Xn is surface functional groups [112, 113, 114]. MXenes possess a wide range of components and structures, and some are competitive with other 2D materials. MXenes are also promising as electrocatalysts due to good hydrophilic surfaces, easily modified structures, excellent electrical conductivities, and superb chemical stability [115, 116]. And as a result, Vojvodic et al.  predicted and validated that Mo2CTx catalysts are active toward HER through experimental results and found that unlike MoS2 in which only edge sites are active, the basal planes of Mo2CTx are also active catalytic sites for HER. In general, the HER performance of MXenes largely depends on oxygen terminations on the surface due to their strong interaction with H intermediates , and Gao et al.  demonstrated that as compared with MXenes without oxygen terminations, those with oxygen terminations possess lower hydrogen absorption free energy, resulting in a high exchange current density for HER. In addition, MXenes have also been found to be active for ORR and OER after suitable modifications [119, 120], allowing these newly developed 2D materials to not only act as catalysts and co-catalysts, but also as supporting substrates in electrocatalysis . Overall, the multifunctionality of MXenes makes them promising candidates for catalytic applications.
BP is another representative 2D material that has attracted increasing attention because of its unique structures and properties [121, 122]. BP nanosheets possess large surface areas, lone-pairs of electrons on the surface, and anisotropic electrical and optical properties, making them promising electrocatalysts for water electrolysis [123, 124]. The first study of BP electrocatalysts was reported by Jiang et al.  in which synthesis was carried out by using a thermal-vaporization transformation method, resulting in an OER performance with an overpotential of 0.37 V to achieve a current density of 10 mA cm−2, comparable to commercial RuO2 electrocatalysts. However, similar to LDHs, the OER performance of bulk BP is also limited by insufficient active sites and the most effective method to improve the performance is the exfoliation of bulk BP to few-layer nanosheets. For example, Ren et al.  developed a simple liquid exfoliation method to obtain few-layer BP nanosheets and reported OER performances that were significantly better than those of bulk BP. In addition, because BP is a new nonmetal layer material, methods used to improve the catalytic performance of graphene and TMDCs may also be applicable. Another issue with BP is that it is unstable in water and its layers gradually cleave and generate defects on the edges and surfaces, which may cause issues for electrocatalytic applications in aqueous media. To address this, Wang et al.  developed an effective method to protect BP through the synthesis of Co2P on the defects/edges of BP nanosheets. Here, the researchers reported that Co2P on the edges not only occupied defect sites to improve stability, but also increased electrocatalytic performances. Overall, the use of BP in electrocatalysis has only been recently studied, and it is expected that the research on BP-based high-performance electrocatalysts will continue to expand.
2D metal nanosheets (e.g., Ru, Ir, or Pt nanosheets) have also been widely studied in electrocatalysis due to their high intrinsic activities, high specific surface areas, good charge transfer properties, and abundant active sites, making them promising as high-performance electrocatalysts [128, 129, 130]. In addition, 2D metal alloy nanosheets are potential substitutes for precious metal-based electrocatalysts because they can reduce the amount of precious metals used [131, 132]. Researchers report that the maximization of atom exposure is necessary to achieve high catalytic activity in 2D metal nanosheets. Based on this, Kong et al.  synthesized free-standing 2D Ru nanosheets with large portions of atom exposure and reported that the resulting 2D Ru nanosheets demonstrated improved HER activities with an overpotential of 20 mV at 10 mA mg−1, which is lower than that of commercial Ru powder (30 mV). Furthermore, Ir-based electrocatalysts have also been widely applied for OER. For example, Pi et al.  investigated 2D Ir nanosheets with large surface areas and accessible active sites and reported greatly improved catalytic activities toward OER in which onset potentials of 1.45 V and 1.43 V versus RHE in acidic and basic media were obtained, respectively. Moreover, catalytic properties of 2D metal nanosheets can be greatly enhanced through the introduction of transition metals to optimize electronic structures and adsorptive behaviors. And overall, these results highlight that the rational construction of 2D structures can enhance the electrocatalytic performance of 2D metal-based catalysts.
3.2 2D Material-Based Heterostructures for Electrocatalysis
Interfacial structures such as heterostructures are key in determining the catalytic performance of 2D materials and can not only overcome the intrinsic shortcomings of each material, but also produce novel properties due to interfacial effects. Based on this, the design and fabrication of heterostructures composed of 0D, 1D, and 2D materials is an effective method to achieve high electrocatalytic performances. In recent years, there has been growing interest in the synthesis of 2D material-based heterostructures to produce high-performance electrocatalysts. The following sections of this review will discuss the fabrication and performance of heterostructures based on 2D materials, including 0D/2D, 1D/2D, and 2D/2D heterostructures.
3.2.1 0D/2D Heterostructures
0D materials are materials possessing sizes in all three dimensions of less than 100 nm and include single atoms, clusters, quantum dots, and nanoparticles. These materials have been widely used as catalysts due to the exposure of abundant active sites with high catalytic activities. However, 0D materials possess low electrical conductivities, making them unsuitable for electrocatalysis if used alone. To resolve this, an effective method is to disperse 0D materials on supports with large surface areas and good conductivity. For example, based on the fact that single-metal atoms can be prepared on high surface area 2D materials to make 0D/2D heterostructures for electrocatalysis, Fei et al.  prepared an HER electrocatalyst based on small numbers of individual cobalt atoms dispersed on NG (Co-NG) in which the resulting Co-NG demonstrated good electrocatalytic performances for hydrogen production with a low overpotential of ~ 170 mV at a current density of 10 mA cm−2. In another example, Cheng et al.  developed a 0D/2D heterostructure composed of individual Pt atoms and clusters on an N-doped graphene nanosheets (Pt/NGNs) in which the size and the dispersion of the Pt atom clusters were accurately controlled through atom layer deposition (ALD) (Fig. 8a). Here, the researchers conducted both experimental and theoretical studies to reveal that Pt atoms preferred to adsorb onto nitrogen sites in NG and possessed strong interactions with NG, thus promoting electron transfer between them. And as a result, Pt/NGNs in this study provided exceptionally high catalytic activities and good stability for HER as compared with commercial Pt/C catalysts (Fig. 8b) as well as the Co-NG electrocatalysts. Other 2D materials such as TMDCs have also been used as supports to improve the catalytic performance of 0D/2D heterostructures. For example, Cheng et al.  used Rh nanoparticles and MoS2 nanosheets, as strong H-adsorbing and quick H2-desorbing components, respectively, in which the resulting Rh/MoS2 heterostructures demonstrated a low overpotential of 47 mV at a current density of 10 mA cm−2, a small Tafel slope of 24 mV dec−1, and decent stability. Here, the excellent performances of the catalysts were attributed to the Rh atoms and their fast capture of hydronium ions as a result of their strong H-adsorbing ability in which the adsorbed H atoms can migrate to the surface of MoS2 with strong H2-desorbing ability, demonstrating the mechanism in which 0D Rh nanoparticles and MoS2 nanosheets together can improve the electrocatalytic activity of heterostructured catalysts. Despite this, the large-scale synthesis of single atom or nanoparticle catalysts dispersed on 2D materials is difficult.
Like metal atom catalysts, 0D compounds can also serve as excellent active sites for electrocatalytic reactions [138, 139, 140, 141]. For example, Li et al.  in situ grew MoS2 nanoparticles onto RGO using a one-step solvothermal method (Fig. 8c–d) and reported that because 0D MoS2 nanoparticles possess more abundant S edge sites than 2D MoS2 nanosheets, their combination with RGO can significantly promote catalytic activities, leading to a small onset potential of ~ 0.1 V (Fig. 8e) and a low Tafel slope of 41 mV dec−1 (Fig. 8f). In addition, the researchers cycled the resulting MoS2/RGO catalyst for 1000 cycles and reported negligible cathodic current losses from the initial value (Fig. 8g), demonstrating good durability. Aside from small onset potentials and overpotentials, large current density is another important parameter for high-performance electrocatalysts that is often ignored. Based on this, our group recently designed and constructed a 0D/2D heterostructure electrocatalyst based on Mo2C nanoparticles and MoS2 nanosheets (MoS2/Mo2C) through the in situ carbonization of MoS2 . Here, the resulting MoS2/Mo2C heterostructure was found to possess highly exposed active sites and rough surfaces both at the micro- and nanoscale and demonstrated a low overpotential of 227 mV in acidic media and 220 mV in alkaline media at a large current density of 1000 mA cm−2. In addition, the MoS2/Mo2C catalyst also provided good durability during a 24-h test in both media. Here, the super performance of this MoS2/Mo2C heterostructure was attributed to the enhanced interfacial mass transfer and the surface oxygen formed on Mo2C during HER. And based on these results, further studies into practical industrial use should be conducted. Overall, 0D/2D heterostructures are promising for electrocatalysis; however, poor long-term durability prevents practical application and further improvements are necessary.
3.2.2 1D/2D Heterostructures
Different from substrate-supported electrodes, self-supporting electrodes are more suitable for practical applications. And to optimize the electrocatalytic interfacial structure of self-supporting electrodes, Chen et al.  fabricated a graphene-CNT hydrogel film electrocatalyst doped with both N and O (NG-CNT) through the layer-by-layer assembly of graphene and CNTs that can be used as a self-supporting electrode for electrocatalysis (Fig. 9h). Here, graphene and CNTs were assembled in a relatively ordered fashion in the hydrogel film due to their strong interactions (Fig. 9i), which the researchers suggested may facilitate charge transport and improve durability during catalytic processes. In addition, the resulting NG-CNT was found to be highly hydrophilic with a small contact angle of ~ 1.6º as compared with 74.2º for dry NG-CNT (Fig. 9j), indicating that the reactants in the electrolyte have easy access to the surface of the film. And as a result, the NG-CNT film directly used as a working electrode for OER provided high catalytic activities with a small onset potential of 315 mV (Fig. 9k), which was less than that of other samples such as dry NG-CNT (410 mV) and G-CNT (322 mV). Furthermore, this NG-CNT hydrogel film with its unique structural features also provided surprisingly high performances that were even better than noble metal oxides (IrO2) and some transition-metal complex catalysts [150, 151, 152, 153]. The researchers in this study also reported that NG-CNT provided similar catalytic currents as scan rates increased from 10 to 100 mV s−1, suggesting highly efficient transports and favorable catalytic kinetics within the electrodes (Fig. 9l). More importantly, NG-CNT also demonstrated excellent durability and stability with little performance loss (Fig. 9m). Overall, self-supporting electrodes based on 1D/2D heterostructures have been demonstrated to possess decent performances and can be directly used in energy conversion devices in which performance improvements will come from better coordination between 1D and 2D materials in terms of nanostructures, chemical components, and the interactions between them.
3.2.3 2D/2D Heterostructures
Similar to 1D/2D heterostructures, 2D/2D heterostructures can also serve as self-supporting electrodes for direct energy conversion devices. For example, Duan et al.  constructed a flexible film by integrating porous C3N4 (PCN) nanosheets with nitrogen-doped graphene (PCN@-N-graphene) using a simple vacuum filtration method in which the large numbers of in-plane and out-of-plane pores in the PCN nanosheets can generate highly exposed active sites. Here, the flexible film with a hierarchical porous structure allowed for enhanced mass transport during catalytic processes and the layer-on-layer structure of C3N4 and graphene possessed strong interactions to enhance charge transfer. And as a result, the self-supporting PCN@-N-graphene electrode produced superior HER performances with a small onset potential (− 0.008 V) that was close to that of commercial Pt, a high exchange current density of 0.43 mA cm−2 and excellent durability with negligible activity losses after 5000 cycles, allowing this 2D/2D heterostructure with good conductivity, flexibility, and catalytic performance to be promising for practical electrocatalysis applications. Beyond 2D/2D heterostructures, 2D materials can also possess electronic coupling with substrates in a face-to-face fashion. For example, Voiry et al.  reported that the electronic coupling between MoS2 and a gold substrate can greatly reduce the contact resistance of systems and improve electron injection from the substrate to the catalyst active sites. Here, the basal plane of 2H MoS2 is generally less active than 1T MoS2 for HER due to its worse conductivities and poorer charge transfer kinetics. Therefore, the facilitation of charge transfer is an effective method to increase the activity of basal planes in 2H MoS2 . And because gold substrates possess excess d electrons, they can improve the charge transfer of 2H MoS2 if the two are coupled together. And as a result, the electrons injected from the gold substrate to the basal plane of 2H MoS2 can not only accelerate charge transfer but also enhance the adsorption of hydrogen reactants onto the basal planes of 2H MoS2, resulting in improved electrocatalytic performances of 2H MoS2. Overall, this study provided new insights into the role of contact resistance and charge transport on the catalytic performance of 2D materials, and the research discussed in this section indicated that the catalytic performance of 2D materials can be substantially improved through the construction of 2D/2D heterostructures, laying the foundation for 2D materials development in electrocatalytic applications.
4 Conclusions and Outlook
The exploration of novel 2D materials with intrinsically high catalytic activities. Here, the control and engineering of both the chemical and structural composition of these 2D materials need to be considered to achieve this goal, and these explorations will not only assist in the understanding of the origins of active sites, but also provide insights into the effects of doping, defects and strain on catalytic properties. In addition, model electrocatalysts based on 2D materials need to be built to enable simple and more reliable theoretical calculations.
The elucidation of electrocatalytic mechanisms of 2D materials with the help of advanced in situ characterization techniques. Electrochemical reactions involve multiple electron transfer processes in which each electron transfer is accompanied by the formation of different intermediates. The identification of these intermediates is key to understanding the catalytic mechanisms. Here, in situ characterization techniques such as in situ X-ray absorption fine structure spectroscopy are helpful in analyzing the structural and valence state of intermediates.
The high-efficiency and energy-saving synthesis of 2D material-based electrocatalysts in large quantity is a prerequisite for practical applications. Currently, most studies only focus on enhancements in catalytic performance and often ignore the costs of individual components and the complexity of the synthesis process. Therefore, the development of efficient strategies for the large-scale production of high-quality 2D materials is urgently needed.
The durability and the stability of 2D material-based electrocatalysts need to be taken into consideration because they are important parameters in the evaluation of electrocatalysts. The durability of an electrocatalyst mainly depends on its chemical stability in solution, its aggregation between layers and its connection with supporting electrodes in which the structure and components of electrocatalysts must remain stable to maintain activity during long catalytic processes.
The electrocatalytic activity of 2D materials based solely on laboratory data is often inaccurate, and it is difficult to fairly compare the catalytic performance of 2D materials with other nanomaterials for any given reaction. Therefore, a standardized evaluation system such as large current density and stability for the analysis of electrocatalytic performance needs to be established to meet practical requirements.
By following these suggestions, the study of 2D materials that started just ten years ago can potentially lead to important practical applications.
We acknowledge the financial support from the National Natural Science Foundation of China (Nos. 51521091 and 51722206), the Youth 1000-Talent Program of China, the Shenzhen Basic Research Project (No. JCYJ20170307140956657), the China Postdoctoral Science Foundation (No. 2018M641346), the Economic, Trade and Information Commission of Shenzhen Municipality for the “2017 Graphene Manufacturing Innovation Center Project” (No. 201901171523), and the Development and Reform Commission of Shenzhen Municipality for the development of “Low-Dimensional Materials and Devices” disciplines.
- 8.Electrocatalysis for the generation and consumption of fuels. Nat. Rev. Chem. 2, 0125 (2018). https://doi.org/10.1038/s41570-018-0125
- 18.Zheng, Y., Jiao, Y., Jaroniec, M., et al.: Advancing the electrochemistry of the hydrogen-evolution reaction through combining experiment and theory. Angew. Chem. Int. Ed. 5, 52–65 (2015)Google Scholar