Imidazolatic-Framework Bimetal Electrocatalysts with a Mixed-Valence Surface Anchored on an rGO Matrix for Oxygen Reduction, Water Splitting, and Dye Degradation.

This paper presents a simple strategy for manufacturing bifunctional electrocatalysts-graphene nanosheets (GNS) coated with an ultrafine NiCo-MOF as nanocomposites (denoted NiCo-MOF@GNS) having a N-doped defect-rich and abundant cavity structure through one-pool treatment of metal-organic frameworks (MOFs). The precursors included N-doped dodecahedron-like graphene nanosheets (GNS), in which the NiCo-MOF was encompassed within the inner cavities of the GNS (NiCo-MOF@GNS) at the end or middle portion of the tubular furnace with several graphene layers. Volatile imidazolate N x species were trapped by the NiCo-MOF nanosheets during the pyrolysis process, simultaneously inserting N atoms into the carbon matrix to achieve the defect-rich porous nanosheets and the abundantly porous cavity structure. With high durability, the as-prepared nanomaterials displayed simultaneously improved performance in the oxygen reduction reaction (ORR), the oxygen evolution reaction (OER), and photocatalysis. In particular, our material NiCo-MOF@GNS-700 exhibited excellent electrocatalytic activity, including a half-wave potential of 0.83 V (E ORR, 1/2), a low operating voltage of 1.53 V (E OER, 10) at 10 mA cm-2, a potential difference (ΔE) of 1.02 V between E OER, 10 and E ORR, 1/2 in 0.1 M KOH, and a low band gap of 2.61 eV. This remarkable behavior was due to the structure of the defect-rich porous carbon nanosheets and the synergistic impact of the NPs in the NiCo-MOF, the N-doped carbon, and NiCo-N x . Furthermore, the hollow structure enhanced the conductivity and stability. This useful archetypal template allows the construction of effective and stable bifunctional electrocatalysts, with potential for practical viability for energy conversion and storage.

Introduction

Widespread efforts have been made to develop renewable energy technologies, including fuel cells, metal–air batteries, water splitting devices, and heterogeneous catalysis, to meet the rapidly expanding demands for clean green energy and environmental remediation.[1−4] The applicability of heterogeneous catalysis is continuously growing, along with demands for new effective catalysts. Such energy storage devices might also require catalysts for the oxygen reduction and oxygen evolution reactions (ORR and OER, respectively) for water division, as well as for photodegradation, to directly enhance their energy conversion and storage efficiencies.[5] Systems for effective water splitting and photocatalysis, however, require semiconductor catalysts that can efficiently drive the water splitting process by absorbing solar light and subsequently generating excitonic charge carriers.[6−8] To achieve successful bifunctional ORR and OER water splitting, as well as dye degradation, a catalyst material should display a desirable band edge position with respect to water oxidation, a low degree of electron–hole recombination, high electrochemical activity in water, and rapid charge transport.[9,10]

Various inexpensive, stable, and earth-abundant electrocatalysts have been tested to date for their merits in ORR and OER catalytic operations; among them, transition-metal oxides (TMOs) are promising substitutes for noble-metal catalysts (e.g., Pt, Au, Pd).[11−14] Most TMOs, however, display poor electronic conductivity, ready aggregation of nanoparticles (NPs), and a low surface area, resulting in unsatisfactory bifunctional catalytic activities; this problem be alleviated by mixing graphene nanosheet (GNS) materials with TMO NPs.[15−17] Because of their excellent physicochemical properties (e.g., high surface area, excellent electrical conductivity, electrochemical stability, and ease of surface modification), nanostructured carbonaceous materials [e.g., GNS, carbon nanotubes (CNTs), carbon nanofibers (CNFs)] have been studied extensively as promising electrocatalyst materials for several electrochemical applications (e.g., water splitting, energy storage, photocatalysis). In particular, the one-dimensional (1D) nanostructures, high aspect ratios, tunable multilayer-structured morphologies, high porosity, light weight, and structural stability of GNS materials have resulted in enhanced electrochemical and photochemical water splitting properties as well as the potential for application in versatile OER water splitting processes.[18] Although some noble-metal-based catalysts, including Ir and Ru, and their corresponding oxide compounds, have recently exhibited high OER activity, their high cost and scarcity have restricted their large-scale use.[19−22]

Bimetallic electrode materials having the structure [M1–2+M3+(OH)2][A·mH2O] (where M2+ and M3+ are divalent and trivalent metal ions, respectively, and A is a charge-balancing anion) are attractive for their potentially large electrochemically active surface, high stability, and cationic brucite-like layer materials (featuring host layers, interlayer charge reimbursing anions, and solvation molecules).[18,23,24] Because of the tunable composition of their host layers, the tunable content of interlayer anions, and the presence of hydroxide ligands extending into the interlayer space, the metallic shape of such materials as OER electrocatalysts has attracted significant attention.[23,24] Furthermore, as a result of their larger surface areas, greater numbers of defects or vacancies, and conductivity higher than that of their bulk counterparts, the bimetallic forms of hydroxide nanostructures have exhibited remarkably superior OER properties.[25,26]

The diversity and designability of the structures and compositions of metal–organic frameworks (MOFs) have attracted increasing attention in the field of materials science and chemistry. The high consistency of MOFs with a unique range of metal ions makes it possible to use them in electrocatalysis applications to build homogeneous multimetallic materials. In addition, the crucial electronic and/or surface structures of hollow and/or porous nanostructures can be tuned with the use of MOFs, thereby promoting electrocatalytic activity.[27−29] For example, Wang et al.[30] prepared Fe–Ni-enclosed layered porous nitrogen-doped graphene (PNG) NPs starting with MOFs as electrocatalysts for the bifunctional OER and hydrogen evolution reaction (HER) without the intervention of an electrocatalytic ink dispersion or drop-casting. A Ni-MOF has been employed by Wang et al. as a prototype and decorated with NiOOH active sites.[30] Their OER catalyst-based Fe-mediated Ni-MOFs were prepared using a hydrolysis strategy. The special three-dimensional (3D) flower-like structure adorned with NiOOH active sites, induced by Fe3+ species, appeared to be responsible for the excellent OER electrocatalytic action. For the degradation of methyl orange (MO) dye, Liu et al.[32] used a hydrothermal process to synthesize flower-like hierarchical Ni–Co layered double hydroxide (Ni–Co LDH) microspheres. The size of microspheres and thickness of the Ni–Co LDH nanosheets are roughly about 5–6 μm and 25 nm in diameter, respectively. While these exploratory studies have revealed the high potential for applying mixed-metal MOFs for the OER, the use of MOFs in electrocatalysis remains limited by their poor conductivity, their small pore size, and their organic ligands embedded active metal centers. In addition, it remains a major challenge to produce highly active transition-metal MOF-based OER catalysts displaying rapid kinetics for four-electron (4e–) processes.[32] More importantly, inadequate knowledge of the catalytic mechanisms is further limiting the optimal design of inexpensive transition-metal MOF-based catalysts for high-efficiency water splitting and dye degradation.

In this study, we used ultrasonication to prepare methanolic dispersed GNS as a conductive layer. In addition, we developed a simple method to manufacture a new hierarchical porous Co–Ni-dependent methylimidazole MOF (NiCo-MOF) with GNS, employing an ultrasonication process and varying the Ni/Co molar ratio. The structural dodecahedron NiCo-MOF, which accumulated on the surface of the GNS, consisted of arrays of nanosheets comprising self-convening dodecahedron structures. The hierarchically arranged dodecahedron NiCo-MOFs provided a large number of electroactive surface sites to promote electron transport and electrolyte ion diffusion in the NiCo-MOF/GNS nanocomposites, thereby facilitating redox reactions while also enhancing the transport of charges by inhibiting the recombination of electron–hole pairs. In 1.0 M KOH, the optimized NiCo-MOF@GNS-700 material provided a remarkably low overpotential (ca. 420 mV at a current density 10 mA cm–2) and a small Tafel slope (ca. 84.8 mV dec–1); its electrochemical activity was superior to that of commercial 20% Pt/C and RuO2 catalysts. We have characterized the crystal structures, microstructures, surface functional groups, and porosities of these new materials. Our study of the catalytic reaction mechanism has revealed that the OER activity and the photocatalytic activity were affected by the contents of Ni/Co and their various chemical states.

Results and Discussion

We used a simple three-step process, including ultrasonic Ni-MOF synthesis and subsequent soaking of Co(NO3)2·6H2O solution, to prepare the NiCo-MOF@GNS electrode materials (Scheme ). The pristine NiCo-MOF powder was first manufactured through ultrasonic treatment of 2-MIM, Ni(NO3)2·6H2O, and Co(NO3)2·6H2O in MeOH at room temperature, and then for 3 h in the presence of the dispersed GNS solution. Imidazolic group as a second aromatic heteroatom source carried out by ultrasound energy to agitate with ZIF-67@Ni-ZIF dodecahedron nanoparticles to doped nitrogen into precursor and form NiCo-N active species and highly porous graphene nanosheets (GNS). During the carbonization process, the metals Ni and Co promote the conversion of the carbon source into thin-layer graphene carbon, which then curls and shrinks to form graphene nanosheets composites with N-containing rich defect and abundant cavity structure trapped in ultrafine NiCo-MOF nanoparticles, which benefits to improve conductivity and stability.

Scheme 1

Schematic Representation of the Fabrication of NiCo-MOF@GNS-Based Nanocomposites for ORR and OER Water Splitting

Because the surface and electrocatalytic properties of carbon-based materials can be varied by adjusting the sintering temperature, the behavior of the electrocatalysts was optimized by subjecting them to different sintering temperatures. The morphologies of the as-prepared NiCo-MOF electrocatalysts that had been sintered at 500, 600, 700, 800, and 900 °C were examined using scanning electron microscopy (SEM). The polyhedral morphology was well preserved after annealing in each case. Furthermore, because of the uniform decoration of NiCo-MOF NPs, the polyhedral surface was rough after annealing, consistent with previous studies.[33] NiCo-MOF NPs have been demonstrated previously to act as effective catalysts when combined with a GNS matrix. SEM images of the NiCo-MOF@GNS sample prepared from the 2-MIM surface, and those that had been subjected to subsequent annealing under an Ar atmosphere, are displayed in (Figure a–e).

Figure 1

(a–e) SEM images of the NiCo-MOFs prepared with sintering at (a) 500, (b) 600, (c) 700, (d) 800, and (e) 900 °C under Ar. (f–h) TEM images (low and high magnification) of the optimized NiCo-MOF@GNS-700 electrocatalyst. (i) SAED pattern of NiCo-MOF@GNS-700.

The NiCo-MOF@GNS sample displayed higher roughness for the polyhedral surface spread over graphene, due to the wrapping of the thick GNS-oriented network. The average length of the NiCo-MOF@GNS particles produced from the 2-MIM hybrid was approximately 2.8 nm, with an average diameter of approximately 4 nm (Figure f). More significantly, the dense growth of directed GNS in the NiCo-MOF@GNS material limited the agglomeration of the GNS and provided promising conductive pathways to accelerate the rates of electron transfer in the ORR and OER and the passage of photoactivated electrons, as well as catalyst surface holes. We expected these features of the NiCo-MOF@GNS composites to result in high-efficiency ORR and OER water splitting and photocatalysis for the degradation of organic dyes. Figure g,h reveals that the GNS was wrapped by the polyhedral surface of the NiCo-MOF during controlled annealing under the Ar atmosphere. The large content of the 1D graphene surface assisted the in situ growth of the NiCo-MOF on the GNS surface through coordinate bonds; transmission electron microscopy (TEM) revealed the formation of large amounts of carbon via the NiCo-MOF catalyst after annealing. For the NiCo-MOF@GNS-700 sample, the interlayer spacing of the GNS was approximately 0.316 nm, corresponding to the (002) plane (Figure h), consistent with the X-ray diffraction (XRD) data (see below). In addition, the coexistence of NiCo-MOF and the GNS (average particle size: ca. 5–15 nm) suggested that the NiCo-MOF@GNS composites would serve as useful electrocatalysts. As displayed in Figure h, the lattice fringes of the NiCo-MOF (ca. 0.316 nm) that surrounded the GNS layers were distributed to the (002) plane of the Co-based metal–organic frameworks. In addition, the selected area electron diffraction (SAED) pattern revealed the presence of the smallest diffraction ring, corresponding to the nature of planes (111), (220), (311), and (400), implying that the materials centered on NiCo were polycrystalline.[34] To induce a more electrically conductive network and promising photocatalytic behavior for the degradation of organic dyes, the presence of the GNS template would presumably be highly advantageous.

To examine whether crystalline phase changes occurred to the products during the carbonization process, we recorded XRD patterns of the Ni–Co-MOFs that had been sintered at 500, 600, 700, 800, and 900 °C (Figure a). All of the XRD patterns featured two peaks located at 43.96 and 51.31° for an extreme composite of NiCo-MOF@GNS, corresponding to the lattice planes (111) and (200) of the Ni and Co components. Furthermore, all XRD patterns featured a large characteristic diffraction peak at 25.81°, matching the graphitic carbon (002) crystal face.[33−38] These features indicated that the GNS carbon additive was distributed onto the skeleton of the NiCo-MOF, creating more active sites and a greater pore volume on the surface of the GNS. The carbon (GNS) contents, evaluated through elemental analysis (EA) of the samples sintered at 500, 600, 700, 800, and 900 °C were approximately 20, 24, 27, 35, and 42 wt %, respectively. We also characterized the NiCo-MOF@GNS-700 sample, which displayed the best OER and ORR performances, using micro-Raman spectroscopy (Figure b).

Figure 2

(a) XRD patterns and (b) Raman spectra of the Ni–Co-MOFs prepared with sintering at 500, 600, 700, 800, and 900 °C under an Ar atmosphere. (c) N2 sorption isotherms and (d) pore size distribution of NiCo-MOF@GNS-700.

Figure b reveals the graphitic carbon structure of the NiCo-MOF@GNS samples prepared with sintering at various temperatures. The D (defect) band appeared at 1330.58 cm–1, representing disordered graphite and a graphitic lattice vibration mode with an A1g symmetry. The G (graphitic) band centered at 1585.3 cm–1 is an ideal graphitic lattice vibration mode with E2g symmetry. The D band represents the topological defects and disorder of carbonaceous materials; further defects in sp2-hybridized carbon materials are represented by the high strength of the D band. In addition, the G band emerges from the sp2-hybridized carbon in-plane vibrations. These characteristic peaks confirmed the graphic structure of our as-prepared NiCo-MOF@GNS nanocomposites. Both the D and G bands of the NiCo-MOF@GNS nanocomposites were extended, also revealing the disorder of the graphite plane. The defect density can be characterized in terms of ID/IG, the ratio of the intensities of the G and D bands, respectively. The ID/IG ratio can be used to identify the degrees of surface/edge defects and disorder, as well as the mean crystalline size of the sp2-hybridized domains in NiCo-MOF@GNS samples derived from MOFs.[34,39−42] A high ID/IG ratio indicates defect-rich positions in nanomaterials, as well as a high degree of graphitization and high conductivity. The ID/IG ratios for the NiCo-MOF@GNS-500, -600, -700, -800, and -900 nanocomposites were 0.921, 1.08, 0.954, 1.01, and 1.08, respectively. The highest ID/IG ratios for NiCo-MOF@GNS-600, -800, and -900 suggested that they featured more defect sites but poor electrical conductivity, whereas for NiCo-MOF@GNS-500 the lower value of ID/IG implied better electrical conductivity but less defect sites. The NiCo-MOF@GNS-700 composite, with its intermediate value of ID/IG, presumably featured more defect sites, a strong degree of graphitization, and high conductivity, encouraging of its use in electrochemical nanomaterial catalysis.

We obtained the Brunauer–Emmett–Teller (BET) detailed surface area and pore size distributions (PSDs) from N2 adsorption/desorption isotherms and nonlocalized density functional theory (NLDFT) to investigate the porous nature of our optimized catalyst composites (Figure c,d). Interestingly, NiCo-MOF@GNS-700 provided a type IV isotherm, with an exponential increase in N2 adsorption at a relatively low pressure (P/P0 < 0.01), suggesting the presence of micropores. In addition, the presence of an H4 hysteresis loop in the NiCo-MOF@GNS-700 hybrid isotherms at relative pressures (P/P0) ranging from 0.50 to 0.98 suggested the presence of medium and large mesopores. For NiCo-MOF@GNS, the measured BET specific surface area was 135.5 m2 g–1, substantially greater than the previous estimate.[43,44]Figure d presents the PSDs of the optimized as-synthesized hybrid catalyst (NiCo-MOF@GNS-700), determined using the NLDFT approach. Interestingly, the PSD plot shows the NiCo-MOF@GNS-700 particle size with a diameter of 3.00 nm for the pore width and 1.29 cm3 g–1 for the pore depth. The high surface area and large graphene-assisted NiCo-MOF hybrid pore volume would be very advantageous for improving the electrochemical properties and for promoting the adsorption of organic dyes on the surface.

For as-prepared NiCo-MOF@GNS-700 (Figures a–d and S1), XPS was used to describe the surface chemical compositions of nanomaterials and the chemical valence of metal ions. NiCo-MOF@GNS composite exhibits XPS survey spectra of Ni 2p, Co 2p, C 1s, O 1s, and N 1s, demonstrating composites containing Ni, Co, C, O, and N components, as seen in XPS patterns. Co 2p spectra can be divided into two pairs of spin–orbit doublets and two shakeup satellites, as seen in high-resolution spectra (787.36 and 805.21 eV). Two pairs of doublets are observed at (780.36, 795.45 eV) and (781.97, 797.71 eV), respectively, which correspond to Co 2p3/2 and Co 2p1/2 of Co0 and Co3+. Co2+ is responsible for the peak at 780.36 eV. Co2+ and Co3+ species coexist due to oxidation of Co nanoparticles on the surface of NiCo-MOF@GNS during the preservation process. Similarly, the four peaks located at 854.71 and 872.78 eV belong to metallic Ni, and the binding energies at 861.64 and 879.85 eV correspond to satellite, as shown in Ni 2p patterns. These findings show that NiCo alloys form and that NiCo alloys redistribute atomic charge between Ni and Co species, facilitating the adsorption of reaction intermediates during ORR, OER, and photoelectrocatalytic processes. Three peaks in the C 1s spectra, at 284.93, 285.72, and 289.15 eV, correspond to the C–C, C–N, and C=O characteristics peaks for NiCo-MOF@GNS composites. The characteristic peaks of NiCo–O, C–O, and oxygen vacancies in O 1s spectra associated with binding energy at 530.25, 532.53, and 533.90 eV, respectively, indicating weakly bound oxygen in hydroxyl groups that chemically adsorbed oxygen or water on the surface of NiCo-MOF@GNS composites. In all samples, five peaks with binding energies of about 399.09, 399.81, 401.44, 402.50, and 406.30 eV were observed in N 2p spectra, named pyridine-N, NiCo-N, pyrrole-N, graphite-N, and oxidation-N, respectively. All of the above findings show that as the pyrolysis temperature rises, the graphite-N content rises, as does the content of other N species, implying that different N species can convert to graphite-N. The existence of graphite-N in NiCo-MOF@GNS composites could improve conductivity, according to a related article. Electroactive materials with a lot of N-doped carbon (especially pyridine-N) can act as active sites and improve ORR catalysis efficiency. NiCo alloy particles and NiCo-N species are favorable for OER and photoelectrochemical reactions. Pyridine-N, which is found at the graphene edge, improves capacitance, and easily combines with metal to form the metal-N-C (MNC, M = CoNi) group as a catalytic active site, demonstrating the synergistic effect of NiCo-N to N-doped carbon in NiCo-MOF@GNS composites to develop electrocatalytic efficiency.[34,39−43] The coexistence of pyridine-N, NiCo-N, and graphite-N is thus regarded as a cluster of catalytic active sites, which benefits ORR, OER, and photoelectrocatalytic activity of NiCo-MOF@GNS composites.

Figure 3

XPS spectra of NiCo-MOF@GNS-700: (a) survey, (b) C 1s, (c) O 1s, and (d) N 1s.

Electrochemical Performance

To evaluate the performance of the as-fabricated NiCo-MOF@GNS materials for the OER and the ORR, we applied the LSV technique after coating on GCEs at a catalyst loading of 3 μg cm–2. The electrocatalytic OER activity of the NiCo-MOF@GNS samples was studied in O2-saturated 0.1 M KOH electrolyte using a standard three-electrode system. Figure a displays the iR-corrected polarization LSV curves of the various electrocatalysts for the OER at a scan rate of 5 mV s–1. The onset potential (ca. 1.53 V) for the NiCo-MOF@GNS-700 sample was significantly lower than those of other catalysts and of the commercial 20 wt % Pt/C and RuO2/C samples; we evaluated the OER activity from the operating potential required to attain a current density of 10 mA cm–2 (E). Remarkably, NiCo-MOF@GNS-700 displayed superior OER activity, with a low overpotential of 420 mV at 10 mA cm–2 and a small Tafel slope (ca. 74.5 mV dec–1), relative to the other NiCo-MOF@GNS (500, 600, 800, 900), while also outperforming the benchmarks of the commercial Pt/C and RuO2 catalysts (η = 570, 520, 470, and 440 mV, and commercial catalysts above 800 and 370 mV, respectively; Tafel slopes of 118.4, 152.4, 212.5, 107.1, 272.5, and 73.4 mV dec–1, respectively), suggesting that an optimization calcination temperature, a hollow microporous structure, and the introduction of a suitable amount GNS carbon additive in the NiCo-MOF were all dynamic factor affecting the OER activity. Furthermore, the Tafel plots and overpotential bar diagram revealed that NiCo-MOF@GNS-700 provided the lowest slope (64.9 mV dec–1) among these tested bimetallic catalysts (Figure b,c), implying that it provided favorable kinetics for the OER. We used electrochemical impedance spectroscopy (EIS) to further examine the OER water splitting kinetics provided by the NiCo-MOF@GNS catalysts that had been subjected to the various calcination temperatures.[41,43] The semicircular diameter of the EIS curves (Rct) provides information regarding the interfacial charge-transfer reactions. Figure d reveals that the charge-transfer resistance (Rct) of NiCo-MOF@GNS-700 (308 Ω) was lower than those of the other composite catalysts that had been subjected to sintering at 500, 600, 800, and 900 °C (3557, 1222, 462, and 647 Ω, respectively), suggesting a more rapid charge-transfer rate that would be beneficial for an efficient OER process.

Figure 4

(a) LSV curves of OER polarization at commercial 20 wt % Pt/C and RuO2 catalysts and at NiCo-MOF@GNS electrodes that had been sintered at various temperatures, recorded using 1 M KOH as the electrolyte at a scan rate of 5 mV s–1 in a N2-saturated medium. (b) Corresponding Tafel plots. (c) Bar diagram comparing the overpotentials at the various catalysts. (d) Nyquist plots recorded at 1.58 V (versus RHE).

To confirm the potential applications of the bifunctional oxygen electrocatalysts, we performed cyclic voltammetry (CV) and LSV experiments in 0.1 M KOH to assess the ORR activity. Figure S2a,b reveals that distinctive oxygen reduction peaks appeared at 0.86 V, but no obvious peaks could be observed as arising from the N2-saturated electrolyte, with superior ORR activity appearing for NiCo-MOF@GNS-700. Figure a indicates that NiCo-MOF@GNS-700 displayed an ORR activity with a more positive onset potential (Eonset) of 0.93 V and half-wave potential (E1/2, ca. 0.83 V), as well as a larger current density (JL), relative to those of the other electrocatalysts (NiCo-MOF@GNS-500, 600, 800, 900, and commercial catalysts RuO2), which displayed performance close to that of the commercial 20 wt % Pt/C. The Tafel plots and bar diagram (Figure b,c) reveal that NiCo-MOF@GNS-700 had a lower overpotential (ca. −290 mV) and smaller Tafel slope (ca. −155.1 mV dec–1) compared with those of the other NiCo-MOF@GNS samples (500, 600, 800, 900), while also outperforming the benchmark commercial catalysts Pt/C and RuO2 (η: −340, −330, −290, −300, −200, and −350 mV, respectively; Tafel slopes: −2411, −216.1, −161.4, −167.1, −199.1, and −235.1 mV dec–1, respectively), suggesting favorable kinetics toward the ORR. The number of electrons (n) involved in the ORR can be calculated from the Koutecky–Levich (KL) eqs and 2 as followsFigure d represents the ORR polarization activities of our as-prepared NiCo-MOF@GNS composite catalysts, recorded at various rotating speeds to assess the electron transfer kinetics. The corresponding KL plots were linear (Figure e,f), providing calculated electron transfer numbers (n) for NiCo-MOF@GNS-500, -600, -700, -800, and -900 of approximately 4.00, 3.89, 3.87, 3.49, and 3.45, respectively, suggesting a four-electron ORR pathway for our NiCo-MOF@GNS catalysts, consistent with that of the commercial Pt/C catalyst. The bifunctional electrocatalytic performance is usually assessed in terms of the difference (ΔE) between the potential of the OER polarization curve at 10.0 mA cm–2 and that of the ORR polarization curve at −2.2 mA cm–2. Figure S4 reveals that NiCo-MOF@GNS-700 provided the lowest value of ΔE (ca. 1.02 V) among our as-prepared catalyst samples as well as the commercial Pt/C. Thus, the NiCo-MOF@GNS-700 composite catalyst exhibited excellent bifunctional catalytic activity toward the ORR and OER, ostensibly because its improved interfacial properties enhanced the adsorption and desorption of O2 at its surfaces and interfaces.

Figure 5

(a) LSV curves of the ORR polarization at commercial 20 wt % Pt/C and RuO2 catalysts and NiCo-MOF@GNS electrodes that had been subjected to sintering at various temperatures, recorded using 0.1 M KOH as the electrolyte at a scan rate of 5 mV s–1 in O2-saturated medium. (b) Corresponding Tafel plots. (c) Bar diagram comparing the overpotentials at the various catalysts. (d) LSV curves for ORR polarization of NiCo-MOF@GNS-700 at optimized temperature and (e) the corresponding KL plots. (f) Bar diagram comparing the number of electrons transferred at various catalysts.

We examined the electrochemical ORR and OER stabilities of our NiCo-MOF@GNS samples, prepared with various calcination temperatures, through double-pulse chronopotentiometry with an O2-saturated 0.1 M KOH electrolyte, alternating between the OER at a current density of 10 mA cm–2 and the ORR at a current density of −1 mA cm–2, with each pulse lasting for 1 h, with the catalyst supported on a GCE, due to reputable adhesion. Figure a reveals that NiCo-MOF@GNS-700 exhibited superior catalytic stability, with an unvarying change in the overvoltage between the OER and ORR after 4 days (ca. 100 h). Figure b presents the multistep chronopotentiometric curve for NiCo-MOF@GNS-700. The current density increases from 2 to 80 mA cm–2, initially during three cycles at 5 mA cm–2, and then it remains steady with increments of 10 mA cm–2 per 60 min. Initially, with an increase of current density, the voltage remained constant for the following 360 min; after 40 mA cm–2, the voltage increased, and a current density of 80 mA cm–2 led to a sudden increase in the voltage, thereafter, remaining constant for the following 480 min. This rapid response reflected the superior mass transport behavior (electrolyte ion diffusion and release of gas products) and high electrical conductivity of the as-prepared NiCo-MOF@GNS-700. Figure S5a–d suggests that low sintering temperatures (500 and 600 °C) resulted in lower carbon contents and a greater concentration of metal ions (Ni2+ and Co2+), leading to poor electrochemical stability; at high sintering temperatures (800 and 900 °C), an increase in the residual carbon content decreased the OER stability because a greater content of amorphous-type carbon products was present on the cathode surface, thereafter suffering from rapid deactivation during the OER cycle, due to physical detachment of the catalyst film caused by the gas bubble effect. Thus, a suitable catalyst and sintering temperature were necessary to achieve long-term stability.[40−44] The properties and electrochemical applications of NiCo-bimetallic nanocomposites previously published in the literature are compared in Table .[11,22,28−32,38] In comparison to the literature, the architecture of our catalyst (dodecahedron nanosheets) is highly desirable and derived from a simple ultrasonication process, which is a faster and more scalable synthesis method. This novel structural design has a high surface area and allows for fast electron transfer. As a result, the NiCo-MOF@GNS-700 electrocatalyst performed better than previously recorded work in terms of low overpotential, ORR, and OER bifunctional stability over 100 h at current densities of −1 and 10 mA cm–2. Chronopotentiometry measurement of the stability of NiCo-LDH@GNS-700 catalysts for ORR and OER water splitting at a current density of −1 and 10 mA cm–2 in O2-saturated 1 M KOH electrolyte solution as shown in Figure S6. From these results, our prepared catalysts NiCo-MOF@GNS-700 have good stability at 20 000 s.

Figure 6

(a) Chronopotentiometry measurement of NiCo-MOF@GNS-700 for ORR and OER water splitting at a current density of −1 and 10 mA cm–2 in O2-saturated 1 M KOH electrolyte solution. (b) Corresponding OER polarization curve revealing a stepwise increase of the current density (no IR compensation of the data).

Table 1

Comparison of the Properties of Our As-Prepared NiCo-Bimetallic Form of Nanocomposite Electrocatalysts with Those of Previously Reported Bifunctional and Other Applications

composites	method	precipitants	applications	stability	references
NiCo alloys	hydrothermal	urea	zinc air batteries		(11)
CoNi-LDH	LBL method	HMT	water oxidation	5.6 h	(22)
NiCo-LDH	soaking method	2-methylimidazole	supercapacitors		(29)
NiCo-LDH	hydrothermal	HMTA	adsorption of MO dye		(28)
N-graphene/CoNi alloy	co-precipitation	urea	fuel cells		(31)
N-CoNi alloy/Janus like carbon	co-precipitation	2-methylimidazole	ORR/OER		(33)
NiCo-LDH	hydrothermal	H3BTC	degradation of organic dye		(38)
MOFs-CoNi alloy-N-doped rich defect	hydrothermal	D-cyanodiamide	ORR/OER/HER	5 h	(30)
dodecahedron NiCo-MOF/graphene nanocomposite	ultrasonication	2-methylimidazole	water splitting and dye degradation	100 h	this work

In brief, we ascribe the superior electrocatalytic performance of NiCo-MOF@GNS-700 for the ORR and OER water splitting to several structural features. First, its hierarchical 3D heterostructure, including the nanoplate morphology, provided a large obtainable surface area for electrolyte ions, potentially facilitating mass transfer and release of the gas bubble product. Second, modifying NiCo-MOF@GNS with additives—in this case, two-dimensional (2D) GNS—not only ensured close contact and better electrical conduction between the components but also generated self-supporting electrodes without polymeric binders (binder-free electrode). Third, the open-ended heterostructure ensured fast electron transfer and sufficient exposure of active sites. Finally, the enhanced electronic conductivity could accelerate the electron transfer kinetics.

Photocatalytic Degradation Performance

To investigate the optical absorbance of the as-prepared catalyst materials, we applied UV–vis diffuse reflectance spectroscopy (DRS, Figure ). As revealed in Figure a, the spectra of the structures prepared with and without modification of the NiCo-MOF and of the bare GNS revealed extended light absorbance capacity into the visible region, potentially facilitating the utilization of solar energy, with absorption edges at 460, 490, and 378 nm, respectively. After modification of the NiCo-MOF by GNS, the optical absorption in the UV–vis region was enhanced. The corresponding band gap energies of these electrocatalysts, calculated using the Kubelka–Munk function (eq ) and Tauc plots (eq ), were as followswhere R is the reflected light percentage, α is the absorption constant, h is Planck’s constant, v is the light frequency, Eg is the band gap energy, A is a constant, and the coefficient n is equal to 1 for the direct transition and equal to 4 for the indirect transition of the photocatalysts. Figure b reveals a predictable optical band gap energy of NiCo-MOF in the metallic form (ca. 2.8 eV) that was consistent with those in previous reports,[47] but slightly modified by the addition of the GNS. For the composite electrocatalyst, the absorbance edges changed, leading to a band gap of approximately 2.61 eV. This small shift in the band gap indicated that the GNS sheets functioned as supports in the NiCo-MOF@GNS nanocomposite, with the formation of a heterostructure with intimate interfacial contact between the carbon and the metallic catalyst. Theoretically, the nano-sized NiCo-MOF@GNS nanocomposite electrocatalyst would deliver a higher contact interfacial area, accelerating charge carrier transport, followed by suppression of charge recombination, resulting in superior photocatalytic activity.

Figure 7

(a) DRS spectra and (b) the corresponding Tauc plots of the as-prepared NiCo-MOF, NiCo-MOF@GNS-700, and bare GNS.

The photocatalytic performance of the as-synthesized NiCo-MOF@GNS-700, NiCo-MOF, and bare GNS were evaluated by investigating the degradation of a rhodamine B (RhB) dye aqueous solution (25 ppm) under natural solar light irradiation. RhB, a basic azo reactive dye, is used extensively industrially as a colorant (Figure S7a,b, Supporting Information). To avoid hazardous effects, it is preferable to eliminate these harmful effluents from the environment using simple and smart technology systems. The UV–vis spectrum of RhB featured a peak at 553 nm, representing the highest occupied molecular orbital–lowest unoccupied molecular orbital (HOMO–LUMO) transition, corresponding to the π–π* transition between the π-systems of the azo groups.[44−46]

Study of Experimental Parameters of Electrocatalysts

To compare the catalytic efficiencies of the bare GNS and the NiCo-MOF@GNS-700 designed photocatalyst, the photodegradation of RhB was performed by dispersing the catalyst (0.05 g) in RhB solution (ca. 25 ppm, 50 mL) and by exposing it to natural solar light for 120 min at different time intervals. Figure a reveals that the absorbance peak of RhB was gradually quenched over time under light radiation, diminishing more predominantly in the case of NiCo-MOF@GNS-700 than for the as-prepared NiCo-MOF or the bare GNS. Interestingly, NiCo-MOF@GNS-700 demonstrated superior activity toward the photodegradation, degrading approximately 94% of the dye after 120 min under solar light. In contrast, the as-prepared NiCo-MOF and bare GNS led to more gradual decreases in the RhB absorbance, with only approximately 88 and 73% of the dye being degraded under similar conditions (Figure a). Because the GNS was decorated with the bimetallic NiCo-MOF, its large surface area could increase the adsorption of dyes onto the photocatalyst, enhancing the photocatalytic degradation activity. To better understand the self-degradation performance of RhB under solar light, we recorded the UV–vis absorbance spectra of RhB in an aqueous medium in the presence of our developed NiCo-MOF@GNS-700 photocatalyst, as a function of the time interval and the pH of the medium (Figure c,d, respectively). A gradual decrease in the intensity of the absorbance band at 554 nm occurred upon increasing the illumination time, with complete degradation of dye obtained after 120 min of illumination.

Figure 8

(a, b) Effect of (a) irradiation time and (b) solution pH on the adsorption of RhB. (c, d) UV–vis absorbance spectra of RhB before and after solar light irradiation in the presence of NiCo-MOF@GNS-700, with variations in (c) irradiation time and (d) pH of the medium.

Adjusting the pH in an examined organic and inorganic system affects the surface charge of an electrocatalyst and the level of dissolved ions, allowing as the effect of the H+ and OH– ions on the uptake sites to be predicted. We explored the influence of pH on the photocatalytic degradation of RhB by changing the pH of an aqueous RhB solution from 3 to 13 at a fixed amount of photocatalyst (0.05 mg mL–1) and dye (25 ppm). Figure b reveals that at a low pH, our as-prepared photocatalysts exhibited poorer degradation efficiency than they did at alkaline pH. The maximum degradation efficiency occurred at pH 11 in the case of the bare GNS and in the case of the NiCo-MOFs with and without modification with the GNS. Therefore, the optimal pH was taken to be 11, allowing the application of these materials to environmental remediation of industrial wastewater having a pH of 10.5–11. The impact of pH on photocatalytic degradation has a complex mechanism and is dependent on several factors, including the ionization state of the photocatalyst, the dye used, and the rate of creation of radicals. As a result, we attributed the higher degradation efficiency at basic pH to the enhanced formation of hydroxyl radicals (•OH), which are strongly oxidizing species. Additionally, at a much higher pH of 13, the generation of •OH radicals would result in increased columbic repulsion between the dye molecules and the negatively charged surface of the photocatalyst, thereby decreasing the degradation efficiency and the penetration of light.

Analysis of the comparative degradation rates revealed that the combination of both materials (94% degradation after 120 min) improved the catalytic efficiency of the individual bare GNS and as-prepared NiCo-MOF (ca. 73 and 88.8% degradation after 120 min, respectively), as displayed in Figure a. We plotted ln(C/C0) with respect to time to investigate the rate of degradation (k). The degradation followed pseudo-first-order kinetics[46] (Figure b), with the values of k indicating the superior catalytic activity for NiCo-MOF@GNS-700 (Table S1, Supporting Information), relative to those of the bare GNS electrode and the pure NiCo-MOF. The higher photocatalytic activity of NiCo-MOF@GNS was due to the presence of the co-catalyst layer of GNS and the bimetallic NiCo; the GNS served as an excellent electron conductor, enabling the movement of electrons through the interfacial layers and preventing electrons and holes from recombining at the same time.

Figure 9

(a) Degradation of RhB in the presence of various catalysts. (b) Pseudo-first-order degradation kinetics in the presence of various catalysts.

The photocatalytic activity of NiCo-MOF@GNS was enhanced by generating the heterostructure using the GNS template. Scheme provides a general overview of the probable NiCo-MOF@GNS photocatalytic process for the degradation of the hazardous dye RhB. Because of its lower band gap energy, the high electrically conductive GNS is very useful for storing and shuttling electrons when coupled with other materials (−4.5 eV). The light-mediated photocatalytic dye degradation essentially involved four steps: (a) adsorption of dye molecules over the surface of the catalyst; (b) light absorption by the photoactive catalyst; (c) generation of electrons and holes in the conduction and valence bands; and (d) creation of highly active radical species, in particular peroxide and •OH radicals, by means of charge-transfer reactions, for dye degradation. The products (CO2 and H2O) are not harmful. After visible light absorption on the surface of the photocatalyst, the photoactivated electrons produced are excited from the filled valence band (HOMO) to the NiCo-MOF LUMO (conduction band). The electron excitation mechanism leads to the creation of an unstable electron/hole pair (eq ) under the irradiation of visible light from graphene to NiCo so that subsequent photogenerated holes require one electron to regenerate stability. Thus, as demonstrated in eq , positive holes in the valence band readily react with the surface-adsorbed H2O to form OH species. Quite the opposite, the electrons in the NiCo-MOF conduction band migrate to a lower Fermi graphene level, resulting in the Schottky barrier on the NiCo-MOF@GNS-700 interface (eq ). Consequently, recombination of the NiCo-MOF electron/hole pair is impeded, and the band gap is narrowed (as evident from UV–vis DRS spectra), promoting the mobility of the charge carrier for RhB degradation.[47,48] In addition, the presence of more GNS tuned the electronic structure, favored the adsorption of more RhB on the elevated surface area of the NiCo-MOF@GNS-700 surface, and promoted the mobility of the charge carrier for dye degradation.

Scheme 2

Projected Mechanism of Photocatalytic Degradation of RhB Dye Mediated by the NiCo-MOF@GNS-700 Photocatalyst under Solar Light

Proposed Mechanism of Photodegradation Activity

To form the superoxide radical anion, hydrogen peroxide, and the •OH radical, electrons in the graphene surface absorb molecular oxygen as per eqs –11. The superoxide and •OH radicals that are produced are highly active oxidants that degrade RhB to form nontoxic products (CO2 and H2O) (eq ) and almost completely remove its color. Therefore, the synergistic effects of graphene, nitrogen doping, NiCo, and GNS combine, leading to the efficient degradation of RhB under visible light—rather than solely the effect of graphene itself, as has been stated in the literature.[48]

Conclusions

In summary, through a simple ultrasonic process, we synthesized a novel N-doped, defect-rich GNS coated with ultrafine NiCo-MOF NPs. Relative to commercial benchmark electrocatalysts, the as-prepared NiCo-MOF@GNS-700 nanocomposite exhibited excellent catalytic efficiency as well as greater stability for the ORR, OER, and photocatalytic reactions. The high catalytic efficiency derived from the defect-rich porous GNS structure and the synergistic effects of the NPs of NiCo-MOF, N-doped carbon, and N-doped NiCo NPs (NiCo-N). Furthermore, the hollow structure of the NiCo-MOF@GNS nanocomposites enhanced the distribution of active sites and substantially improved their stability; in addition, the ultrathin nanosheets were beneficial in shortening the mode of transmission of electrons to increase their electrical conductivity. Our NiCo-MOF@GNS-700 material displayed outstanding ORR, OER, and photocatalytic action, characterized by a small value of ΔE (1.02 V)—the difference between EOER, 10 and EORR, 1/2—in 0.1 M KOH and a low band gap, relative to those of the MOF without modification and of the bare GNS, while also mediating the superior photocatalytic degradation of the RhB dye. This methodology should be useful when extended to the synthesis of other advanced multifunctional composite electrocatalysts as possible substitutes for precious metal materials and applied to various renewable energy conversion and storage technologies.

Experimental Section

Materials

All precursors were of analytical grade and used without further cleaning. 2-Methylimidazole (2-MIM), anhydrous methanol (MeOH), and potassium hydroxide (KOH) were purchased from Sigma-Aldrich; nickel nitrate hexahydrate [Ni(NO3)2·6H2O, 99.0%] and commercial Pt/C (20 wt %) and RuO2 electrocatalysts were obtained from Alfa Aesar; and cobalt nitrate hexahydrate [Co(NO3)2·6H2O, 99.0%] was procured from J. T. Baker Chemicals.

Fabrication of NiCo-MOF@GNS Electrocatalysts

Electrocatalysts comprising NiCo-MOF and GNS were prepared using a simple sonication process. First, multilayer reduced graphene oxide nanosheet (UBIC, Taiwan) powder (2 wt %) was dispersed in MeOH (50 mL) through probe sonication for 1 h at room temperature. Initially, 2 mmol of Ni(NO3)2·6H2O and 4 mmol of Co(NO3)2·6H2O were dissolved in MeOH (80 mL) under sonication until a homogeneous solution had formed. 2-MIM (24 mmol) was added slowly into the homogeneous mixture and then the dispersed graphene oxide was added dropwise slowly into the same mixture with continuous sonication. After 3 h, a blackish-purple color product NiCo-MOF was formed. This product was centrifuged and washed several times with MeOH until the filtrate was colorless, and then dried in an oven at 80 °C for 12 h. Finally, the product NiCo-MOF@GNS nanocomposite was sintered in a quartz tube furnace at 500, 600, 700, 800, or 900 °C under an atmosphere of Ar for 2 h at the heating rate of 2 °C min–1.

Electrocatalysts Characterization

The microstructure and crystallinity of the as-prepared NiCo-MOF@GNS material with different calcined temperatures were analyzed using an X-ray diffraction (XRD) spectrometer (D2 Phaser, Bruker, Germany) with a Cu Kα radiation source. The surface morphology was investigated using scanning electron microscopy (SEM; S-2600, Hitachi, Japan) and transmission electron microscopy (TEM: JEOL JEM-2100, Japan) to examine the porous nature of NiCo-MOF@GNS. Micro-Raman spectral analysis was performed using a confocal micro-Raman spectrometer with 632 nm He–Ne laser excitation (Renishaw RM1000). The carbon residual content in the composite catalyst samples was measured using an elemental analyzer (Thermal Flash 2000). Brunauer–Emmett–Teller (BET) surface areas were determined from N2 adsorption/desorption measurements at 77 K (Tristar-3000, Micromeritics). X-ray photoelectron spectroscopy (XPS) was conducted using an X-ray photoelectron spectrometer (VG Scientific ESCALAB 250, U.K.) and XPSPEAK41 software. UV–visible absorbance spectra were recorded on a PerkinElmer Lambda 35 spectrophotometer.

Electrochemical experiments were performed using a CHI 405 workstation and a three-electrode system. The water splitting tests were performed in 0.1 M KOH. To obtain an electrocatalyst ink, NiCo-MOF GNS (10 mg) was dispersed in water (0.4 mL), isopropyl alcohol (IPA, 0.550 mL), and Nafion solution (ca. 5–6%, 0.05 mL) under sonication. A glassy carbon electrode (GCE; diameter: 3 mm) was polished and coated with the as-prepared electrocatalyst ink as the working electrode. The reference and counter electrodes were Pt wire and silver/silver chloride (Ag/AgCl), respectively. The suspension of a well-dispersed electrocatalyst (12 μL) was placed on the GCE surface and left to dry for a few minutes at 60 °C in an oven (catalyst loading: ca. 0.22 mg cm–2). All of the currents displayed herein are faradic currents after correction for the capacitive effect. The potential calculated versus an Ag/AgCl electrode was transformed, according to eq to a potential versus a reversible hydrogen electrode (RHE). After repeated cycling (10 times) between the ORR and OER, all data were collected. To evaluate the operation of the ORR/OER, the relationship between current and overpotential at an electrical interface is given by the well-known Butler–Volmer eq , as followswhere I is the current, I0 is the exchange current (current at equilibrium potential), αA and αC are the charge-transfer coefficients for the anodic and cathodic reactions, respectively, n is the number of electrons transferred, F is the Faraday constant (96 485 C), R is the ideal gas constant, T is the absolute temperature in K, and η is the overpotential. Under equilibrium conditions, the contribution of both anodic and cathodic terms is equal where the observed current I is equal to I0. Additionally, the high overpotential approximation of the above Butler–Volmer equation will lead to the Tafel eqs and 16 of the anodic and cathodic polarizations, respectively, given below.Equations and 16 have the form of y = a + bmx and will give a linear plot when log I is plotted against versus η, which is the well-known “Tafel plot”.[49]