Hydrogen (H 2 ) has emerged as a promising alternative to fossil fuels, due to its high energy density and environmental cleanliness [1,2]. To date, hydrogen production relies primarily on coal gasification or steam methane reforming, both of which are energy-intensive and non-sustainable, and generate substantial carbon dioxide, a green-house gas [3]. Electrochemical water splitting has emerged as a viable and clean alternative technology for green hydrogen production [4]. Yet appropriate catalysts are required to boost the electron-transfer kinetics of the hydrogen evolution reaction (HER) at the cathode, where platinum-based catalysts represent the commercial benchmarks [5]. However, the high cost and low abundance of platinum significantly hinder the applications. It is thus of fundamental and technological significance to develop low-cost, high-efficiency alternative catalysts [6]. Towards this end, transition metal dichalcogenides (TMDs), such as MoS 2 , WS 2 , CoS 2 and CoSe 2 , have been examined as viable HER catalysts thanks to their unique electronic structure and abundance of tunable active sites [7,8]. In particular, CoS 2 and CoSe 2 demonstrate a relatively low energy for H adsorption on both the Co cations and S/Se anions, favorable electronic configuration and inherently good electrical conductivity, making them highly advantageous as HER electrocatalysts [9,10]. For example, Li et al. [11] synthesized needle-like Co 3 O 4 @CoS 2 core-shell nanostructures on carbon cloth by thermally treating Co(OH) (CO 3 ) ½ ⋅xH 2 O in a H 2 S atmosphere in a tube furnace and observed an apparent HER activity with an overpotential (η 10 ) of -152 mV to reach the current density of 10 mA cm -2 in 0.5 M H 2 SO 4 , due to enhanced electrical conductivity by the unique core-shell structure and formation of abundant active sites at the Co 3 O 4 /CoS 2 interface. In another study, Dai et al. [12] prepared CoSe 2 hollow microspheres loaded on graphene oxide (GO) via a hydrothermal method, achieving an η 10 of -210 mV in 0.5 M H 2 SO 4 , which was ascribed to the unique coupling effects between the structural components of CoSe 2 and GO.
The HER activity can be further enhanced with the formation of cobalt chalcogenide heterostructures, due to the synergistic effect of multiple phases, enhanced electronic interactions, increased exposure of active sites and optimized coordination environment [13,14]. For instance, Liu et al. [15] prepared CoS 2x Se 2(1-x) nanowire arrays supported on carbon fibers with a tube furnace and observed a low η 10 of -129.5 mV in 0.5 M H 2 SO 4 , thanks to optimal engineering of the electronic structure. Karuppasamy et al. [16] synthesized carbon-based CoS 2 and CoSe nanocomposites by a hydrothermal procedure and observed an η 10 of -164 mV in 0.5 M H 2 SO 4 , due to the formation of heterostructures and synergistic electronic interactions between CoS 2 and CoSe. Sun et al. [17] prepared two-dimensional CoSe 2 /Co 3 S 4 heterostructured nanosheets on Co 3 O 4 nanofibers and observed an η 10 of -165 mV in 1 M KOH, which was accounted for by the increased number of active sites and more favorable H adsorption.
In these previous studies, the catalysts were synthesized predominantly using the conventional tube furnace and hydrothermal methods, both of which are energy-and time-intensive. Recently, magnetic induction heating (MIH) has emerged as a highly effective technique for the ultrafast preparation of various functional nanocomposites for HER and OER (oxygen evolution reaction) electrocatalysis, thanks to rapid Joule's heating within seconds that facilitates the formation of abundant nonequilibrium structures [18]. This method has been successfully popularized in a range of materials, such as FeNi spinel oxides [19], amorphous MoS x nanocomposites [20], RuCl x -modified Ru nanoparticles [21], Co nanoparticles encapsulated within defective carbon [22], and RuCu nanocomposites [23]. In this study, nanocomposites with CoS 2 /CoSe 2 heterostructures supported on N-doped carbon (CoS 2 /CoSe 2 /NC) were prepared via a two-step procedure that entailed the derivation of cobalt/N-doped carbon (Co/NC) composites by pyrolysis of cobalt-containing zeolitic imidazolate framework-67 (ZIF-67) followed by MIH treatment of Co/NC at controlled induction currents for 10 s with the addition of elemental S and/or Se. Among the series, the CoS 2 /CoSe 2 /NC nanocomposite with a 1:1 S:Se atomic ratio exhibited the best HER performance in acidic media, with a low η 10 of -187 mV and Tafel slope of 40.7 mV dec -1 , along with superb durability and stability. The performance was significantly enhanced in comparison to those of the monocomponent counterparts of CoS 2 /NC and CoSe 2 /NC, due to the deliberately engineered heterostructure and synergistic effect that manipulated the electronic structure, enhanced the electrical conductivity, and optimized the coordination environment, as confirmed by microscopic and spectroscopic measurements and further validated by theoretical studies based on density functional theory (DFT) calculations.
2-Methylimidazole (CH 3 C 3 H 2 N 2 H, 99%, Acros Organics), cobaltous nitrate hexahydrate (Co(NO 3 ) 2 ⋅6H 2 O, Reagent Grade, Thermo Fisher Scientific), sulfur (S, colloidal, Sigma-Aldrich), selenium (Se, 99.7%, Thermo Fisher Scientific), potassium hydroxide (KOH, certified ACS, Fisher Chemicals), Pt/C (20 wt%, Alfa Aesar), methanol (CH 3 OH, certified ACS, Fisher Chemicals), carbon disulfide (CS 2 , certified ACS, Thermo Fisher Scientific), Nafion 117 solution (95%, Aldrich), 2-propanol ((CH 3 ) 2 CHOH, HPLC Grade, Fisher Chemicals), and sulfuric acid (H 2 SO 4 , 98%, Fisher Chemicals) were used as received. Deionized water was acquired from a Barnstead Nanopure water system (resistivity 18.3 MΩ cm).
ZIF-67 was prepared by following a literature procedure [24]. In brief, 1.092 g of Co(NO 3 ) 2 ⋅6H 2 O was dissolved in 30 mL of methanol and transferred to a flask containing 1.232 g of 2-methylimidazole in 30 mL of methanol. The resulting mixture was sonicated for 15 min and subsequently stirred at room temperature for 24 h. The purple precipitate was collected by centrifugation at 6000 rpm for 5 min, rinsed three times with methanol, and dried under vacuum at 60 • C overnight, yielding the ZIF-67 powders.
The Co/NC nanocomposite was prepared by controlled pyrolysis of ZIF-67. In Brief, the ZIF-67 powders obtained above were loaded onto a ceramic boat, which was then placed into a tube furnace and calcined at 900 • C for 3 h at a heating rate of 5 • C min -1 under a flowing N 2 atmosphere, affording the Co/NC black powders.
The CoS 2 /CoSe 2 /NC nanocomposites were prepared by adopting an MIH procedure reported previously [18][19][20][21][22][23]. Briefly, 25 mg of the obtained Co/NC powders were grounded in a mortar with 75 mg of a mixture of sulfur and selenium powders at a controlled S:Se feed ratio. The well-mixed powders were then transferred onto a piece of graphite paper (2.5 cm × 2.5 cm × 0.01 mm) sitting on top of a same-size iron sheet (2.5 cm × 2.5 cm × 0.2 mm). The assembly was centered on a firebrick inside a quartz tube, purged with Ar gas for 10 min, and then placed into a four-turn induction coil (5 cm in diameter). MIH treatment was conducted at a controlled current of 200 A for 10 s. The resulting sample was denoted as CoS 2 /CoSe 2 /NC. CoS 2 /NC and CoSe 2 /NC were prepared in the same manner but with the addition of 75 mg of sulfur or selenium powders, respectively. A control sample was also prepared at the same S and Se feed but in a tube furnace in a nitrogen atmosphere at 450 • C for 30 min (at a heating rate of 5 • C min -1 ), which was denoted as CoS 2 /CoSe 2 /NC-TF.
Transmission electron microscopy (TEM) measurements and elemental mapping analysis based on energy-dispersive X-ray spectroscopy (EDS) were performed with a JEOL 2100-F operated at 200 kV. Xray diffraction (XRD) patterns were acquired with a Bruker D8 Advance diffractometer using Cu K α radiation (λ = 0.15418 nm). Raman measurements were carried out using a Horiba Jobin Yvon LabRAM ARAMIS automated scanning confocal Raman microscope under 532 nm excitation. X-ray photoelectron spectroscopy (XPS) studies were performed with a Thermo Fisher K-alpha system with the binding energy calibrated against that of the C 1s electrons (284.6 eV). Nitrogen sorption isotherms were collected with a Micromeritics Tristar 3020 Porosimeter system at 77.3 K. X-ray absorption spectroscopy (XAS) measurements were carried out at 10 K at beamline 4-1 of the Stanford Synchrotron Radiation Lightsource using an Oxford liquid helium cryostat.
Electrochemical measurements were carried out using a CHI 710 electrochemical workstation in a three-electrode configuration, consisting of a Ag/AgCl reference electrode, a graphite rod counter electrode, and a glassy carbon rotating disk working electrode (surface area 0.196 cm 2 ). The reference electrode was calibrated against a reversible hydrogen electrode (RHE) and all potentials in this study were referenced to the RHE. The catalyst inks were prepared by dispersing 2 mg of the samples prepared above into a mixed solution containing 740 μL of isopropanol, 250 μL of H 2 O and 10 μL of Nafion under sonication for 30 min. 20 μL of the inks was dropcast onto the surface of the glassy carbon disk electrode and dried naturally at room temperature before being coated with 5 μL of 20 wt% Nafion, corresponding to a catalyst loading of 0.204 mg cm -2 . The HER polarization curves were collected in 0.5 M H 2 SO 4 at the scan rate of 10 mV s -1 and rotation rate of 1600 rpm with 90% iR compensation. Chronopotentiometric measurements were conducted at the current density of 10 mA cm -2 for 30,000 s. Electrochemical impedance spectroscopy (EIS) measurements were conducted within the frequency range of 0.1-10 5 Hz at an AC amplitude of 5 mV using a Gamry Reference 75 instrument, and the data were fitted by using the Randles equivalent circuit.
First-principles computations were carried out using Quantum ESPRESSO, an open-source plane-wave code [25], with the models of CoS 2 (111), CoSe 2 (211), and CoS 2 /CoSe 2 heterostructure (211). A cutoff of 40 and 240 Ry for kinetics and charge density was chosen with the ultrasoft pseudopotential [26]. The smearing parameter was set at 0.01 Ry in the Marzari-Vanderbilt smearing for all calculations [27]. For geometric relaxation, the convergence was 10 -8 Ry of the electronic energy and 10 -4 au for the total force. Density functional perturbation theory was performed to calculate the phonon frequency as inputs for entropy and zero point energy [28].
The preparation of CoS 2 /CoSe 2 /NC nanocomposites consisted of two major steps (Fig. 1a). First, Co/NC composites were prepared by controlled pyrolysis of ZIF-67 in a tube furnace at 900 • C for 3 h. Next, the obtained Co/NC was thoroughly mixed with a controlled amount of sulfur and selenium powders, and converted into CoS 2 /CoSe 2 /NC nanocomposites via MIH treatment at 200 A for 10 s. The MIH process leverages the Joule's heating effect, capable of generating a temperature over 1000 • C within seconds [18][19][20][21][22][23].
The materials structures were first characterized by TEM measurements. From Fig. 1b and c, the CoS 2 /CoSe 2 /NC sample can be seen to exhibit an irregular morphology, a marked deviation from the dodecahedral structure of the original ZIF-67 precursor [22]. In high-resolution TEM measurements (Fig. 1d), the sample can be seen to possess two sets of well-defined lattice fringes, one with interplanar spacings of 0.272 and 0.319 nm that can be assigned to the (200) and (111) planes of CoS 2 (PDF 41-1471) [29,30], and the other with the interplanar spacings of 0.207 and 0.241 nm corresponding to the (220) and (211) planes of CoSe 2 (PDF 09-0234), respectively [31,32]. Furthermore, one can see that these CoS 2 and CoSe 2 nanocrystalline domains were in intimate contact with a distinct grain boundary, confirming the successful formation of CoS 2 /CoSe 2 heterostructures. In EDS-based elemental mapping analysis (Fig. 1e), the elements of C, N, O, Co, S and Se can be clearly resolved and uniformly distributed throughout the sample, in good agreement with the formation of CoS 2 and CoSe 2 species within the carbon scaffold. In addition, the atomic content was estimated to be ca. 34.68 wt% for Co, 11.27 wt% for S, and 16.08 wt% for Se (Table S1), corresponding to an atomic ratio of Co:S:Se = 1:0.60:0.35. The marked deviation from the 1:2 stoichiometric ratio between Co and S/Se suggests that sulfurization and selenization of Co/NC was incomplete and confined within the cobalt surface layers (Fig. 1a).
The materials structures were further probed by XRD measurements. From Fig. S1, one can see that CoS 2 /NC possessed a series of diffraction peaks at 2θ = 32.3
• , 36.2 • , 39.8 • , 46.3 • , and 55.0 • , corresponding to the (200), (210), (211), (220), and (311) planes of CoS 2 (PDF 41-1471), respectively [33,34]. For CoSe 2 /NC, the characteristic diffraction peaks can be resolved at 2θ = 30.6 • , 34.3 • , 37.6 • , 43.6 • , 51.7 • , 56.7 • , 58.9 • , 73.9 • , and 76.1 • , due to the (200), ( 210), ( 211), ( 220), (311), ( 230), (321), (421), and (332) planes of CoSe 2 (PDF 09-0234), respectively [35]. Notably, these diffraction patterns remained rather visible in CoS 2 /CoSe 2 /NC, consistent with the formation of a mixed lattice [36].
Nitrogen sorption measurements were then performed to investigate the specific surface areas and pore size distributions. As shown in Fig. 2a, Co/NC exhibited a Type IV isotherm with a hysteresis loop, indicative of the predominant formation of mesopores [37]. Upon the introduction of S and/or Se, the resulting samples exhibited a Type III isotherm instead, likely because the formation of CoS 2 and/or CoSe 2 diminished the porosity of the samples and hence adsorption of nitrogen onto the carbon scaffold [38]. In fact, one can see that the specific surface area dramatically decreased from 135.00 m 2 g -1 for Co/NC to 11.77 m 2 g -1 for CoS 2 /NC, 2.79 m 2 g -1 for CoS 2 /CoSe 2 /NC, and 1.51 m 2 g -1 for CoSe 2 /NC (Table S2). The pore volume also diminished accordingly from 0.163 cm 3 g -1 for Co/NC to 0.044 cm 3 g -1 for CoS 2 /NC, 0.019 cm 3 g -1 for CoS 2 /CoSe 2 /NC, and 0.008 cm 3 g -1 for CoSe 2 /NC (Fig. 2b, Table S2). This is consistent with the collapse of the dodecahedral structure of the ZIF-67 skeletons during the sulfurization/selenization process (Fig. 1b and e).
XPS measurements were then carried out to investigate the elemental composition and valence states of the samples. Fig. S2 shows the survey spectra of the sample series, where the Co 2p, O 1s, N 1s, C 1s, S 2p/Se 3p, and Co 3p/Se 3d electrons can be readily resolved at ca. 782, 531, 400, 285, 164, and 56 eV, respectively. The corresponding elemental contents are listed in Table S3-S4, where the S:Se atomic ratio was estimated to be ca. 1:1 for CoS 2 /CoSe 2 /NC (this is somewhat different from the results from EDS measurements (Table S1), likely because XPS is more surface-sensitive). Fig. 3a shows the high-resolution scans of the Co 2p electrons. One can see that CoS 2 /NC possesses a doublet at 782.14/797.93 eV due to the 2p 3/2 /2p 1/2 electrons of Co 2+ species, with the corresponding satellite peaks resolved at 785.76/801.55 [39,40]. For the CoS 2 /CoSe 2 /NC and CoSe 2 /NC samples, the Co 2+ 2p 3/2 /2p 1/2 doublet can be resolved at lower binding energies, 781.40/797.19 eV for the former and 781.44/797.18 eV for the latter, suggesting electron enrichment by elemental Se during MIH treatment. In fact, a minor doublet can also be resolved at ca. 779/794 eV for these two samples that can be ascribed to metallic cobalt (Table S5) [41].
From the S 2p spectra in Fig. 3b, it can be seen that CoS 2 /NC consisted of two doublets at 163.61/164.80 and 168.98/170.14 eV that can be assigned to the 2p 3/2 /2p 1/2 electrons of S 2- 2 and SO 2- 4 , respectively (Table S6) [42,43]. For CoS 2 /CoSe 2 /NC, these 2p 3/2 /2p 1/2 electrons redshifted markedly to 162.57/163.68 eV and 168.45/169.61 eV. Meanwhile, the content of SO 2- 4 diminished significantly, and S 2- 2 became the dominant species, again, in good agreement with the Se-driven electron enrichment, as observed above. The Se 3d spectra are shown in Fig. 3c. One can see that the Se 2- 2 3d 5/2 /3d 3/2 electrons can be identified at ca. 55.25/55.98 eV for CoS 2 /CoSe 2 /NC and 55.34/56.05 eV for CoSe 2 /NC (Table S7), along with a small amount of SeO 2 at ca. 60 eV [44,45].
Fig. S3a exhibits the C 1s spectra of the series of samples, where three components, sp 2 C, sp 3 C, and oxidized C, can be deconvoluted at ca. 284.60 eV, 285.34 eV, and 286.89 eV, respectively. This suggests efficient carbonization of the ZIF-67 precursor [46]. In fact, in Raman measurements, the samples all exhibited a well-defined D band (1345 cm -1 ) and G band (1582 cm -1 ), with a rather low band intensity ratio (I D /I G = 0.49 to 0.60, Fig. S4), suggesting a high degree of graphitization during pyrolytic conversion of ZIF-67 to Co/NC [47,48].
The corresponding N 1s spectra are shown in Fig. S3b, where deconvolution yields three N species at 398.95 eV for pyridinic N, 400.23 eV for pyrrolic N, and 401.46 eV for graphitic N, confirming effective N doping within the carbon matrix [49]. The O 1s scans are displayed in Fig. S3c, where the C=O and C-O peaks are identified at ca. 531.94 eV and 533.29 eV, respectively [50]. The fact that no O species can be resolved below 530 eV suggests the absence of metal oxides in the samples. These results are consistent with the formation of CoS 2 /Co-Se 2 /NC heterostructures within the nanocomposites.
The valence band maximum (VBM) profiles of the samples are shown in Fig. 3d. It can be seen that both CoS 2 /NC and CoSe 2 /NC possessed a positive VBM of +1.79 and + 0.57 eV, respectively, suggestive of the semiconducting nature of the materials; by contrast, the VBM became negative at -0.07 eV for CoS 2 /CoSe 2 /NC, indicating a conductor-like behavior most likely due to the formation of CoS 2 /CoSe 2 heterojunction [51]. Furthermore, the fact that the VBM of CoS 2 /CoSe 2 /NC was the closest to the Fermi level among the samples signifies the highest activity in electron-transfer reactions (vide infra) [52].
XAS measurements were then carried out to gain further insights into the sample structure. From the Co K-edge X-ray absorption near-edge structure (XANES) spectra in Fig. 3e, one can see that all samples exhibited a pre-edge feature at ca. 7010 eV, due to the 1s→3d transition of the octahedral symmetry, in contrast to the Co foil reference [53]. The first derivatives of the XANES profiles are shown in Fig. S5, where one can see that the absorption edge energy of the sample series varied in the order of Co foil < CoSe 2 /NC < CoS 2 /CoSe 2 /NC ≈ CoS 2 /NC. This variation of the Co valence state is in good accord with results from the above XPS measurements.
The corresponding Fourier transforms of the extended X-ray absorption fine structure (FT-EXAFS) are displayed in Fig. 3f. CoS 2 /NC and CoS 2 /CoSe 2 /NC can be seen exhibiting a main peak at ca. 1.86 Å due to the Co-S path and another one at ca. 3.52 Å to the Co-Co path. For CoSe 2 /NC the Co-Se path can be resolved at ca. 2.10 Å (which also coincides with the Co-Co path in Co foil) [54]. The peak positions align well with those for pristine CoS 2 and CoSe 2 , confirming the successful formation of CoS 2 and CoSe 2 in the samples.
The fittings of the Co K-edge FT-EXAFS spectra are depicted in Fig. S6 and the results are listed in Table S8. Notably, the bond lengths of Co-S (2.29 Å) and Co-Se (2.40 Å) were very close among the sample series, but slightly shorter than those of pristine CoS 2 (2.32 Å) and CoSe 2 (2.45 Å) [54]. Concurrently, the Co-Co path length enlarged somewhat in the order of pristine CoS 2 (3.91 Å) < CoS 2 /NC (3.92 Å) < CoS 2 /CoSe 2 /NC (3.94 Å), suggestive of a slightly increasingly defective lattice with the formation of CoS 2 /CoSe 2 heterostructures. This is consistent with the slight decrease of the coordination number (CN) of Co-S, pristine CoS 2 (6) > CoS 2 /NC (4.84) > CoS 2 /CoSe 2 /NC (3.93). Meanwhile, the CN of Co-Se can be seen to also decrease from 2.84 for CoSe 2 /NC to 0.77 for CoS 2 /CoSe 2 /NC, all markedly lower than that of pristine CoSe 2 (6), suggesting the formation of largely amorphous CoSe 2 in the samples [55,56].
Notably, the resulting CoS 2 /CoSe 2 /NC nanocomposites exhibited a significantly improved HER performance in acidic media, as compared to CoS 2 /NC and CoSe 2 /NC. Fig. 4a displays the HER polarization curves in 0.5 M H 2 SO 4 , and CoS 2 /CoSe 2 /NC can be seen to stand out as the best catalyst among the series of samples, with the lowest η 10 of -187 mV, as compared to -400 mV for Co/NC, -321 mV for CoS 2 /NC, -257 mV for CoSe 2 /NC [57]. Moreover, CoS 2 /CoSe 2 /NC exhibited the lowest slope in the corresponding Tafel plot (Fig. 4b), in the order of CoS 2 /CoSe 2 /NC (40.7 mV dec -1 ) < CoSe 2 /NC (73.2 mV dec -1 ) < CoS 2 /NC (139.1 mV dec -1 ) < Co/NC (193.5 mV dec -1 ). Additionally, as depicted in Fig. 4c, CoS 2 /CoSe 2 /NC displayed a decay of only 5 mV for the η 10 after 5000 potential cycles within the potential range of -0.25 to -0.20 V, suggesting good stability of the sample for prolonged operation. In chronopotentiometric measurements (Fig. 4d), CoS 2 /CoSe 2 /NC also demonstrated superior stability with only a 13 mV day after a continuous operation for 30,000 s at the current density of 10 mA cm -2 , as compared to 19 mV for commercial Pt/C. The remarkable stability of CoS 2 /CoSe 2 /NC was also manifested in prolonged stability tests for up to 100 h (Fig. S7). Consistently, from the Nyquist plots in Fig. 4e, CoS 2 /CoSe 2 /NC possessed the lowest charge-transfer resistance (R ct = 26.3 Ω at the overpotential of -200 mV), in comparison to CoS 2 /NC (296.2 Ω) and CoSe 2 /NC (157.2 Ω) (Table S9), indicating the fastest electron-transfer kinetics for HER. Taken together, the results suggest that CoS 2 /CoSe 2 /NC stood out as the best HER catalyst among the series. CoS 2 /CoSe 2 /NC even outperformed a number of leading cobalt dichalcogenide catalysts (Table S10), such as CoS 2 -CoSe 2 /CNFs-5 (η 10 = -189 mV) [58], CoSe 2 /SDGC-60 (-203 mV) [59], and CoSe 2 @CNW (-210 mV) [60].
Notably, the HER performance remained virtually invariant for three batches of CoS 2 /CoSe 2 /NC prepared under the same experimental conditions (Fig. S8), indicating good reproducibility of the synthetic procedure (Fig. 1a).
Finally, the samples possessed a very close double layer capacitance (from 4.72 to 5.03 mF cm -2 ) (Fig. 4f), suggesting a similar electrochemically active surface area. This observation, in conjunction with results from the nitrogen sorption measurements (Fig. 2, Table S2), suggests that the enhanced HER performance of CoS 2 /CoSe 2 /NC, as compared to those of CoS 2 /NC and CoSe 2 /NC, was not due to an increase of the accessible surface area, but rather to the heterostructures that enhanced the electrical conductivity and H adsorption, along with the synergistic effect between CoS 2 and CoSe 2 [15][16][17].
In comparison to other samples prepared at different S/Se ratio and MIH currents (Figs. S9-S10 and Table S11-S12), 200 A and a 1:1 S/Se atomic ratio were identified as the optimal conditions for the preparation of the CoS 2 /CoSe 2 /NC heterojunction catalysts. Furthermore, the HER performance was markedly better than that of CoS 2 /CoSe 2 /NC-TF prepared by conventional pyrolysis with a tube furnace (Fig. S11), likely due to the slow heating in tube furnace that rendered the formation of large crystalline domains and hence diminished CoS 2 /CoSe 2 heterojunctions (Fig. S12), despite similar elemental contents and valence states as determined by XPS analysis (Fig. S13, Table S3-S7). This further confirms the unique advantage of MIH in the preparation of nonequilibrium structures [18].
Previously it has been shown that in CoS 2 /CoSe 2 heterostructured composites, the Co, S and Se sites can all contribute to the HER activity; yet the anionic species have been generally argued to be the main active sites (with Se being more active than S), in comparison to the cationic Co species, as manifested by the Gibbs free energy of H adsorption (ΔG H* ), a key descriptor for HER electrocatalysis [57]. In addition, upon the doping of S into CoSe 2 , the |ΔG H* | of both S and Se sites becomes closer to zero [57], and a 1:1 S:Se atomic ratio is preferred for optimal HER performance, due to favorable kinetic migration of H and reduced kinetic energy barrier [61]. Thus, DFT calculations were conducted to unravel the mechanistic insights into the enhanced electrocatalytic activity of CoS 2 /CoSe 2 /NC as compared to CoS 2 /NC and CoSe 2 /NC. Based on experimental characterizations of the materials structures (e.g., Fig. 1d), four structural models for H adsorption were constructed and depicted in Fig. 5a-d
In this study, composites consisting of CoS 2 /CoSe 2 heterostructures supported on N-doped carbon were prepared by rapid MIH treatment. Electrochemical measurements showed that the obtained CoS 2 /CoSe 2 / NC composites exhibited significantly enhanced HER activity in acidic media, in comparison to the monocomponent counterparts of CoS 2 /NC and CoSe 2 /NC. This was ascribed to the heterostructures that led to electron enrichment of the cationic and anionic sites and hence enhanced electrical conductivity and H adsorption, as confirmed by both experimental measurements and theoretical studies. Results from this work highlight the unique potential of MIH in the ultrafast synthesis of transition metal dichalcogenide heterostructures as high-performance electrocatalysts.
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.