Magnetic nanoparticles@Metal-organic framework composites as sustainable electrocatalyst for oxygen evolution reactions

3.1. Chemical and morphological characterization MILs

Thermogravimetric analysis (TGA) with N₂ gas was used to determine the MOF content of the MIL/Fe₃O₄ composite. The TGA analysis (Figure S1(a) & Supporting Information, S3) identified three important weightlessness steps [20]. Figure S1(b) displays the X-ray diffraction (XRD) patterns of Fe₃O₄ magnetite, Fe₃O₄@MIL-53(Fe), and their hybrid structures. The diffraction lines at 2θ > 30° are the typical indices of Fe₃O₄ cubic spinel structure, see Supporting Information, S3, [21].

Fourier transform infrared (FTIR) spectra of MILs samples have been depicted in Fig. S1.c and are similar to each other. The characteristic peaks of a metal-oxo bond between the carboxylic group (Fe-O vibration) of terephthalic acid and Fe (III) have been observed around 578 cm^-1. The peak at 753 cm^-1 for the C-H bending vibration on the benzene cycle and around 1680 cm^-1 vibrations of C=O show the existence of benzene in all samples. Figure S2 (a-c) shows different morphologies of magnetite-based MIL-53(Fe) MOFs. A well-defined bipyramidal hexagonal prism shape was designed for Fe₃O₄@MIL-53(Fe) and La-doped La/Fe₃O₄@MIL-53(Fe) [22]. Ga-doped MIL in Figure S2(c) was providing a Rod-like Fe-based Gd/Fe₃O₄@MIL-53(Fe) [18] The percentage (%) of elements shown in Figure S2 is from EDX analysis.

The brunauer-emmett-teller (BET) specific surface area and pore volume of composites were thoroughly determined by N₂ adsorption-desorption analysis, and the results have been presented in Table 1 and Supporting Information, S3. The low specific surface area of MILs (Table 1), owing to their interpenetrated structure, makes MILs flexible and imparts a large breathing amplitude [23]. The low surface area observed in magnetite-based MIL-53 (Fe) MOFs shows it has almost closed pores with no accessible porosity to N₂ gas. This confirms their ability to adapt their pore opening to accommodate the guest species. This flexibility occurs through tilting of organic linkers without significantly altering the coordination environment around the cation or through a variety of mechanisms, such as rotation of the host framework between a metal ion and an organic ligand [23].

3.2. Catalysts for the oxygen evolution reaction

The task involves evaluating the OER activity on formed MIL samples through electrochemical measurements in 1.0 mol∙L⁻¹ KOH. The electrode is stabilized at a steady-state open circuit potential. Then, cyclic voltammetry (CV) scans are recorded between 0.6 and 1.4 V at a scan rate of 100 mV/s (Figure 1a). In comparison, the La-doped magnetite (La/Fe₃O₄@MIL-53(Fe)) shows a higher oxidation current (∼59.10 mA cm⁻²) than the other two samples (Table 1). This oxidation current is better than Co-MOF/Ti₃C₂Tx@Ni electrode in 3.0 mol∙L⁻¹ KOH [24] but lower than CuCl/CuO(Mn)-NF in 1.0 mol∙L⁻¹ KOH [25], WO₃ NRs@Co-MOF composite catalyst in KOH solution [26].

Close

Figure 1.

Electrochemical OER characterization in 1.0 M KOH, (a) CV at a scan rate of 100 mV∙s^-1; (b) Tafel plots with current densities normalized by the electrochemical active surface area and derived Tafel slopes; (c) the proposed OER reaction pathway on catalyst; (d) Nyquist plots of EIS measurements obtained for the OER at 0.8 V; (e) The equivalent circuit used for the OER EIS data fitting.

Export to PPT

Tafel plots in Figure 1(b) exhibit well-defined Tafel behavior. Differences in Tafel slopes can indicate variations in the OER rate-limiting processes and have previously been linked to the conductivity of the catalysts [27]. The most OER mechanism step in a basic electrolyte is (reactions 1-6) [26-28].

Adsorption of hydroxide (OH⁻(aq)) onto an active site (*) on the anode surface

A second OH⁻ attack leads to deprotonation of *OH and formation of *O, releasing H₂O and an electron (Deprotonation to form adsorbed oxygen)

Another OH⁻ attacks *O, producing *OOH and releasing an electron (Formation of hydroperoxide intermediate (OOH*))

OH⁻ reacts further with *OOH to create peroxo (*OO⁻) and H₂O (Formation of peroxo species and O₂ release)

Then the peroxo species releases O₂ and an electron, regenerating the free active site

The pathway follows, Figure 1(c):

The four-step adsorbate evolution mechanism (AEM) and its intermediates are well-described in reviews on OER, detailing each proton-electron transfer step [29]. These steps indicate that OER begins when a hydroxide ion, OH⁻_(aq), adsorbs onto an active site on the anode. The nanostructured MOF arrangements with saw-tooth morphology may assist the uniformly dispersed M-OOHv in attaining the adsorption of OH ions at the surface of the electrocatalyst, which intensifies the conductance of the material and exhibits OER’s potent electrochemical performance. As in the case of anti-Ferromagnetic RuO₂ [30] and bimetallic Cu–IrO₂ electrocatalyst as OER catalyst, the active site of the catalyst is Ru, Cuand Ir sites on the surface [31]: in our case, the Ln and Fe are the active sites of the studied catalyst. Subsequently, a series of intermediate reactions produces the adsorbed species $*O$ (peroxide), $*OOH$, and $*O{O}^{-}$ (peroxo), culminating in the release of oxygen gas, O₂ (g), from the absorbed species $*O{O}^{-}$. In more detail, Fe₃O₄ is a mixed-valence iron oxide (Fe₂⁺/Fe₃⁺) that provides redox-active sites for OER and Ln (typically as Ln₃⁺) when incorporated, introducing electron-donating characteristics that can alter the electronic density of Fe atoms. The interaction between Ln (La or Gd) and Fe₃O₄ can modulate the d-band center of Fe, improving its binding with oxygen intermediates (*OH, *O, *OOH), which is critical for enhancing OER kinetics, Figure 1(c) [32].

The current density of GCE/La/Fe₃O₄@MIL-53(Fe) is noticeably higher (4.97 mA∙cm⁻²), indicating significant variations in the intrinsic catalytic activity of MILs. The catalyst sample in question had a Tafel slope that was noticeably low (46.82 mV∙dec⁻¹). Comparing this value to those of 63.5 mV∙dec⁻¹ reported by Almarashi et al. for silicon nanowires (SiNWs), [33] 30 mV∙dec⁻¹ reported by Tsai et al. for CFeCoNiP, [34], 30 mV∙dec⁻¹ reported by Heijden et al. for Ni₈₀Fe₂₀OOH, [35] 55 mV∙dec⁻¹ reported for Ni-NiO@3DHPG, 65 mV∙dec⁻¹ reported for o-CoO@3DHPG by Rasheed et al. [36], 55 mV∙dec⁻¹ for r Cu-IrO₂ reported by Ngo et al [31] and 72 mV∙dec⁻¹ for CoNi-MOF compared, 90 and 85 for Co-MOF and Ni-MO, respectively reported by Konavarapuet et al. [37], the low Tafel slope value, suggesting favorable OER kinetics. Suggests a facile electron transfer reaction in bimetallic MOF due to the interaction between the two heteroatoms La⁺³ and Fe⁺ and Fe⁺³, redistributing the electronic density of the active sites, thus facilitating the adsorption of intermediates and the redox process.

It is crucial to remember that nonkinetic effects, such as bubble formation and internal OH gradients, frequently complicate Tafel slope analysis for OER, resulting in apparent Tafel slopes with unclear kinetic significance [35]. The Tafel slope data in Table 1 demonstrate the excellent performance of our prepared electrodes.

Based on the adsorbate evolution mechanism (AEM), Fuhrer et al. [38] stated that a Tafel slope of 40 mV∙dec⁻¹ is most likely associated with the formation of the $*OOH$ intermediate as the rate-limiting step, while the formation of the O intermediate corresponds to a Tafel slope of 64 mV∙dec⁻¹. Tafel slopes ranging from 46.82 to 48.77 mV∙dec⁻¹ obtained in this study may suggest that the formation of either the $\text{*}O$, $*OOH$, or $*O{O}^{-}$ intermediates in steps 2, 3, and 4, respectively, is the rate-determining step (RDS) [39]. The most likely RDS is probably the formation of peroxide in step 3 (FeOOH, LnOOH) (Ln; lanthanide metal). In this scenario, all oxygen atoms evolve from the electrolyte and TMs, giving rise to a higher concentration of surface oxygen vacancies (V_o) and a faster oxygen ion diffusion coefficient, resulting in a smaller Tafel slope. It is important to consider the values of the BET-specific surface area and pore volume of GCE/La/Fe₃O₄@MIL-53(Fe) in Table S1 [17], which contribute significantly to this process.

Additionally, Badreldin et al. demonstrated that the observed volcano-plot pattern for various TM-oxides (La–Ni, La–Mn, La–Cr) indicates that the optimized behavior of TMs is associated with the occupancy of the e_g orbitals (i.e., dz², dx²–y²) in surface TM cations, as well as the occupancy of the 3d electrons [2]. Their findings also reveal that the peak of the volcano plot for La-oxide exhibits an e_g occupancy close to unity, reflecting a high covalency in TM-oxygen bonds, which corresponds to optimal OER activity.

Suntivich et al. contend that in instances where the e_g occupancy is less than unity, the formation of the peroxide ion from the preceding oxyhydroxide represents the RDS. Conversely, if the e_g occupancy exceeds unity, the formation of O–O bonds in the OOH* OER intermediate is identified as the RDS [40].

To further investigate the impact of Fe₃O₄@MIL-53(Fe), La/Fe₃O₄@MIL-53(Fe), and Gd/Fe₃O₄ @MIL-53(Fe) on the OER kinetics, we conducted electrochemical impedance spectroscopy (EIS) test using a sinusoidal voltage of 10 mV per decade and a frequency range of 100 kHz–0.01 Hz, at a potential of 0.8 V vs. Ag/AgCl. The OER Nyquist plots in Figure 1(d) display two adjacent semicircles: the first semicircle is attributed to the porous structure, which arises from the resistance (R_f) of the ionic conducting paths in the pores filled with electrolytes. The second semicircle in the OER Nyquist plot is associated with charge-transfer limitations (R_ct) at the electrode/electrolyte interface. Figure 1(e) illustrates the equivalent electrical circuit used to fit the EIS spectra. The EIS fitting results (Table 2) indicate a slight decrease in the resistance of the first charge-transfer loop for different catalysts, which contributes to an increase in the overall OER rate. The MIL with La-doped magnetite cluster exhibits a notably lower charge-transfer resistance under OER conditions, a behavior that aligns with the findings from the CV and Tafel plot analysis.

Electrocatalytic activity and the electrochemically active surface area (ECSA) are directly correlated. The ECSA was determined through the double-layer capacitance (C_dl), which was calculated using the formula C_dl = j/v, where j is the capacitive current and v is the sweep rate. After estimating the charging current (i_a) based on the CV curves at various scan rates (Figure 2), the double-layer capacitance (C_dl) was calculated for each catalyst. The bipyramidal hexagonal prism structure of GCE/La/Fe₃O₄ @MIL-53(Fe) is expected to yield a high surface area with numerous electrocatalytic active sites. The values of C_dl and ECSA are 65.04.3 mF∙cm⁻² and 3252, respectively. These values are significantly higher than those of the other two compounds: for (GCE/Fe₃O₄@MIL-53 (Fe), C_dl is 39.53.3 mF∙cm⁻² and ECSA is 1976 cm₂; for GCE /Gd/Fe₃O₄@MIL-53 (Fe), C_dl is 40.41 mF∙cm⁻² and ECSA is 2020 cm₂, respectively.

Close

Figure 2.

(a-c) CV curves at the scan rates of 10-300 mV∙s^-1 for MILs different catalysts, The inset double layer capacitance (Cdl) measurements derived from scan rates dependence of current densities; (d) Arrhenius plots of the kinetic currents at 0.8 V; (e) Illustration of different activation energies of OER for Magnetic (blue ball), Gd-doped (Red ball) and La-doped (black ball).

Export to PPT

The kinetic barriers involved in the OER process are represented by the activation energy (Ea). By tracking the current densities from polarization curves at different temperatures, activation energies can be obtained using the Arrhenius equation and transition state theory, as demonstrated by the Ea approach to assess oxygen evolution performance on MIL catalysts. Figure 2(d) shows that the calculated Ea value for the lanthanum-doped magnetic MOF is 8.77 kJ∙mol⁻¹ for GCE/La/Fe₃O₄@MIL-53(Fe), which is significantly better than the 21.0 kJ∙mol⁻¹ for GCE/Fe₃O₄@MIL-53(Fe) and 11.80 kJ∙mol⁻¹ for GCE/Gd/Fe₃O₄@MIL-53(Fe). Comparing these E_a values with some recently developed lanthanide electrocatalysts used for OER (of which there are very few) in Table 3 [41-43], the data demonstrate once again how rare metal-bearing MOFs have become highly promising options due to their enormous surface area, exceptional electrical conductivity, and remarkable electrocatalytic response. These attributes are particularly beneficial for the water-splitting process in fuel cells. Additionally, the layered and highly flexible structure of MOFs provides localized oxygen vacancy channels in the lanthanide plane and a vast number of surface oxygen defects. These structural configurations enhance catalytic activity for OER by significantly lowering the activation energy for oxygen diffusion. Bimetallic MOFs exhibit low overpotential, which is particularly advantageous for the OER process due to the development of active sites created by unsaturated coordination sites, atomic defects, and the diversified valences of the two metals [2].