Transition Metal Chalcogenides for the Electrocatalysis of Water

Sustainable energy technology has received enormous attention in recent years. Specifically, electrochemical water splitting is considered to be the cleanest technique for the production of promising fuels, for example, hydrogen and oxygen, where transition metal (di)chalcogenides (TMCs) as electrocatalysts for hydrogen evolution reaction (HER) and oxygen evolution reaction (OER) have been a grow-ing interest. In this chapter, the typical preparation methods of TMCs such as chemical vapor phase deposition (CVD) and solvothermal synthesis are introduced. Then, several TMC materials for catalyzing HER and OER are reviewed. Most importantly, this chapter also introduced some in situ approaches to realize the mechanism of electrocatalytic behavior toward HER and OER. Finally, the conclusion and futuristic prospects of TMCs in HER and OER are discussed.


Introduction
Energy crisis and environmental pollution arising from the burning of carbonbased fossil fuel in the past decades facilitate people to reconsider the way we utilized the resource on earth. Hydrogen as an ideal energy source came up to the stage due to its high energy density and environmental benignity [1,2]. Electrochemical water splitting is not only regarded as the cleanest technique for hydrogen generation but also suitable to perform on a large scale. The appropriate electrocatalysts are developed to boost the cathodic hydrogen evolution reaction (HER) and anodic oxygen evolution reaction (OER) because of the sluggish kinetics of these two core reactions [overpotential (η), Figure 1a]. Although platinum-based and iridium/ruthenium-based catalysts, respectively, have shown very promising performance in HER and OER, the industrial application is restricted due to their high cost and limited availability [3,4]. Thus, developing alternative electrocatalysts based on non-noble metals or earth-abundant elements is still highly demanded.
The realization of HER and OER mechanisms plays a crucial role to design the efficient electrocatalysts, so the brief discussions on their mechanisms were introduced.
Two kinds of HER mechanisms have been completely studied and widely accepted [5,6]. The first step, a proton from the solution adsorbs onto the catalytic sites of the electrode with a reduction process, is called as Volmer step: Then, either the recombination of two adsorbed hydrogen atoms on the electrode surface is called as Tafel step: or the reaction of an adsorbed hydrogen atom with the hydrated proton, which proceeds with an electron transfer from the electrode surface, is called as Heyrovsky step: Nørskov et al. proposed a Volcano curve showing that experimental exchange currents of materials are as a function of the Gibbs free energy of the adsorbed hydrogen (ΔG H ) (Figure 1b) [7,8]. Ideally, the interaction of hydrogen with the electrode surface should be thermoneutral (ΔG H ffi 0), otherwise either Heyrovsky or Tafel step (strong bonding) or Volmer step would become the rate-determining step. In the case of the OER, many researchers have proposed possible mechanisms at the anode in acidic electrolyte (Eqs. (4)-(8)). However, there are some differences around the reaction of forming oxygen. One route of forming oxygen is through the direct recombination of two MO (M represents Mn, Fe, Co, and Ni metals) (Eq. (6)), while the other route of forming oxygen is through the decomposition of the MOOH intermediate (Eq. (8)).

CVD synthesis of TMCs
The preparation methods of TMCs layers can be categorized into two main approaches: top-down and bottom-up methods (Figure 2). The TMC layers (a) Polarization curves for HER and OER [9] and (b) volcano plot of exchange current density (j 0 ) as a function of DFT-calculated Gibbs free energy (ΔG H ) of adsorbed atomic hydrogen on pure metals [7]. prepared from their bulk material by mechanical exfoliation are called the topdown method, while the TMC layers were produced from the elemental precursors on the target substrate in the bottom-up method, such as chemical vapor deposition (CVD) approach. The CVD method for TMC synthesis has been widely used due to its unique advantages such as scalable size, high crystallinity, and controllable thickness of TMCs [10][11][12][13]. One of the most classic TMCs, molybdenum sulfide (MoS 2 ), has been developed via CVD synthesis to replace the zero bandgap graphene. [14] There are two routes to synthesize MoS 2 layer by CVD (Figure 3a). One is a two-step growth route, where the Mo-based precursors are first deposited and then conducted the sulfurization or decomposition process to form MoS 2 (route 1, Figure 3a). The other growth way of MoS 2 layer is an one-step growth, where the gaseous Mo-and S-based precursors are simultaneously introduced and react to form MoS 2 on a substrate (route 2, Figure 3a). Recent reports have also demonstrated to grow MoS 2 on other kinds of insulating substrates such as quartz, mica, and sapphire (Figure 3b) [15,16]. Figure 3c shows a typical experimental setup for the low-pressure CVD synthesis of MoS 2 . The formation of entire MoS 2 would be prevented according to the ternary ModOdS phase diagram (Figure 3d) if the reducing atmosphere is too weak [17]. Figure 3e displays two possible mechanisms for the growth of MoS 2 . The well-established CVD synthesis of MoS 2 is regarded as a prototype for the synthesis of other TMCs such as MoSe 2 and WS 2 [18,19]. Except from the VI B group metals such as molybdenum and tungsten, the research works of TMCs based on VIII B group metals such as cobalt [20] and nickel [21] are also widely studied in recent years.

Solvothermal synthesis of TMCs
Wet chemical synthesis, a bottom-up process, is broadly used to synthesize TMC nanosheets with tunable thickness and size [24][25][26]. The desired nanosheets form directly in solution or onto substrate in high yield. The synthesis of TMCs via wet chemical method usually relies on the chemical reaction of metal salts and sulfur/selenide-based materials as precursors. The solvothermal method is one of the most typical routes of wet chemical method, in which the reaction is conducted in a sealed autoclave at suitable temperature (Figure 4) [27]. As a representative example, Xie et al. [28] synthesized oxygen incorporated MoS 2 nanosheets from the  precursors of (NH 4 ) 6 Mo 7 O 24 Á4H 2 O and thiourea through a solvothermal method. By tuning the synthetic temperatures, the disordered structure and intrinsic conductivity of MoS 2 could be controlled to a moderate degree. Thus, the optimal catalyst for electrocatalytic hydrogen evolution reaction exhibits onset overpotential as low as 120 mV. It is well known that the materials with different crystal orientations have the anisotropic properties. This phenomenon appears in TMCs depending on stacking sequence of the chalcogen and transition-metal coordination. The commonly found polymorphs in TMC materials are the so-defined 1 T, 2H, and 3R, where the number and letter, respectively, indicate the layer number in the unit cell and the type of symmetry with T, H, and R representing for tetragonal, hexagonal, and rhombohedral, respectively ( Figure 5). For example, Zhang et al. synthesized metallic CoS 2 nanopyramid array on carbon fiber paper by an one-step solvothermal synthesis [29]. In brief, the carbon fiber papers were transferred to the precursor solution containing CoCl 2 Á6H 2 O and CS(NH 2 ) 2 , which was maintained at 180°C for 12 hours. The unique 3D nanostructure and intrinsic metallic properties of the CoS 2 by this method contribute to the ultrahigh activity toward hydrogen evolution reaction.

Hydrogen evolution reaction (HER) based on TMC electrodes
In order to improve HER performance, three main factors including the number of active sites, intrinsic catalytic activity, and the conductivity of TMCs play crucial roles. In 2005, Hinnemann et al. suggested the active sites for HER only exist in the edges of TMCs by density functional theory (DFT) calculation [30]. Thus, many research works are dedicated to enrich the defect sites and/or active sites by nanostructural engineering [31][32][33]. Zhang et al. synthesized the edge-rich 3D MoS 2 coupling with conductive polymer polyaniline (PANI) as catalyst (Figure 6) [34]. The MoS 2 grown on 3D PANI substrate tends to grow vertically and expose abundant edge sites for HER. Consequently, excellent HER performance can be achieved with a low onset potential of 100 mV and a small Tafel slop of 45 mV dec À1 . [34]. Additionally, MoS 2 /PANI achieved superior stability for HER electrocatalysis. Although the promising catalytic activity of MoS 2 for HER was achieved by creating abundant edge sites and/or active sites, the performance was still limited by its intrinsic properties such as poor electrical transport and inefficient electrical contact to the catalyst [35]. Lukowski et al. reported that the metallic nanosheets of 1T-MoS 2 , which were chemically exfoliated by lithium intercalation of semiconducting 2H-MoS 2 nanostructures, dramatically enhanced HER performance (Figure 7) [36]. The current density of 10 mA cm À1 can be reached at a low overpotential of À187 mV with a Tafel slope of 43 mV dec À1 . The excellent performance can be attributed to the favorable kinetics, metallic conductivity, and increasing number of active sites in the metallic 1T-MoS 2 nanosheets, which was proven by the dramatically decrease of charge-transfer resistance from 232 Ω of 2H-MoS 2 nanostructures to 4 Ω of metallic 1T-MoS 2 nanosheets. This finding proves that the metallic 1T polymorph of TMCs is competitive to earth-abundant catalysts in heterogeneous catalysis. Lee et al. developed earth-abundant nanostructuring beaded stream-like cobalt diselenide (CoSe 2 ) nanoneedles (CoSe 2 -BSND) as electrocatalyst for HER [37]. The CoSe 2 nanoneedles derived from the cobalt oxide (Co 3 O 4 ) nanoneedle array directly formed on flexible titanium foils after selenization treatment (Figure 8). The CoSe 2 -BSND can drive the HER at a current density of 20 mA cm À2 with a small overpotential of 125 mV. Also, it possesses a small Tafel slope of 48.5 mV dec À1 suggesting that the HER follows the Volmer-Heyrovsky mechanism where a fast discharge of protons is followed by rate-determining electrochemical desorption. Moreover, the CoSe 2 -BSND electrode achieved great stability in an acidic electrolyte for 3000 cycles. The enhanced electrochemical activity is attributed to the highly accessible surface active sites, the improved charge transfer kinetics, and the super hydrophilic surface of CoSe 2 -BSND electrode.

Oxygen evolution reaction (OER) based on TMC electrodes
The TMC materials have shown promising performance toward HER as we mentioned above. Researchers are searching for the possibility of bifunctional electrocatalysts for both HER and OER to perform the overall water-splitting reaction. Moreover, TMC electrocatalysts have attracted tremendous attentions since Alonso-vante and coworker discovered that Mo 4 Ru 2 Se 8 had a ORR activity comparable to platinum [38]. As compared to ruthenium (Ru) and Rhenium (Rh), the low cost and earth abundant transition metals such as iron-, nickel-and cobalt-based TMCs have much attention for OER [39][40][41]. Liu and coworkers discovered that electrodeposited CoS nanosheet films on Ti mesh show high activity toward OER (Figure 9) [42]. The CoS nanosheets tend to drive a current density of 10 mA cm À2 with an overpotential of 361 mV. In addition, this electrode maintains highly catalytic activity for at least 20 hours. The superior catalytic activity along with excellent stability of CoS nanosheets offers the good opportunity to become a costeffective and industry-feasible electrode toward OER. On the other hand, Swesi et al. first reported that the catalytic activity of OER was observed by the nickel selenide (Ni 3 Se 2 ) in alkaline condition (Figure 10) [43]. The low overpotential  [42]. required to reach 10 mA cm À2 was 290 mV, suggesting that this catalyst exhibits its competitivity among the oxide-based electrocatalysts. The catalytic ability of Ni 3 Se 2 can be further improved through the modification of Se-deficient phase in Ni 3 Se 2 . Moreover, electrodeposited Ni 3 Se 2 catalysts exhibited exceptional stability under OER for 42 hours. The effect of the underlying substrates such as glassy carbon, ITO-coated glass, and Ni foam on OER was also investigated. The results revealed that the glassy carbon substrate exhibited the lowest onset potential and the highest   KOH solution, and (d) in situ Co K-edge XANES spectra of different P-doped CoSe 2 electrodes for HER and OER processes [3]. current density, suggesting that the interaction between the underlying substrate and the Ni 3 Se 2 may play a role in O 2 evolution reaction.
Although the electrocatalytic performance of TMCs toward OER was significantly enhanced, the reaction mechanism and actual active sites responsible for the reaction were in dispute. Recently, operando or in situ experiments like in situ Fourier transform infrared spectroscopy, in situ Raman spectroscopy, and in situ X-ray absorption/diffraction are commonly carried out to provide atomic-level information [3,44,45]. Zhu and coworkers [3] conducted in situ X-ray absorption spectroscopy, in situ liquid-phase TEM, and in situ Raman spectroscopy, revealing that P-doped CoSe 2 in an alkaline solution was acting as the "pre-catalyst" rather than the real reactive species, which has been debated for a while (Figure 11). They found that the introduction of phosphorus would generate more vacancies, which facilitated the structural transformation into the real active electrocatalyst, such as metallic cobalt for HER and cobalt oxyhydroxide (CoOOH) for OER. CoSe 1.26 P 1. 42 shows the best OER performance among all catalysts, which requires an overpotential of 255 mV to reach the current density of 10 mA cm À2 . Furthermore, the CoSe 1.26 P 1.42 catalyst with a Tafel slop of 87 mV dec À1 exhibits a slightly lower than those of the other P-doped CoSe 2 catalysts. Such performance is comparable to many leading earth-abundant HER and OER catalysts in alkaline electrolyte.

Conclusion
Electrocatalysts for HER and OER play an important role for sustainable energy, which require converting renewable energy to storable chemical fuels or employing clean energy. In this article, we have partially reviewed the promising candidates, transition metal (di)chalcogenides (TMCs), from preparation methods to electrochemical measurements toward HER and OER. It can be concluded that a good electrocatalyst should possess good conductivity and moderate adsorption energy to reactive species and/or intermediates. In addition, the large number of active/defect sites would be favorable for catalytic ability. The future challenges for TMCs toward HER and OER are summarized as follows. The realization of the mechanisms toward the HER and OER plays a key role to design a perfect electrocatalyst. The development of in situ or operando techniques shines the light on this difficulty. Most importantly, the understanding of HER and OER mechanisms offers the significant information to many other fields due to the similar operating concepts in electrocatalytic and photocatalytic applications.