Rongmei
Zhu
*a,
Yi
Zhang
a,
Limei
Liu
a,
Yong
Li
a,
Gaihua
He
b and
Huan
Pang
*a
aSchool of Chemistry and Chemical Engineering, Institute for Innovative Materials and Energy, Yangzhou University, Yangzhou, 225009, Jiangsu, P. R. China. E-mail: rmzhu@yzu.edu.cn; huanpangchem@hotmail.com; panghuan@yzu.edu.cn
bSchool of Pharmaceutical Sciences, Jinzhou Medical University, Jinzhou 121001, P.R. China
First published on 2nd July 2024
Metal hydroxides and oxyhydroxides are efficient catalysts for electrochemical oxygen evolution reactions. Herein, we employed a Co-MOF with a tunable structure, high porosity and easy preparation as a precursor to synthesize a bimetal-doped oxyhydroxide/hydroxide electrocatalyst by sequential electrochemical-Lewis acid co-etching and electrosorption doping. This unique co-etching method successfully introduced the high-valent metal ion Hf4+, as well as the electrosorption efficiently doped Fe3+, into the catalyst. Experimental studies and theoretical simulations indicate that the introduction of Hf4+ optimized the OER kinetics, and the introduction of Fe3+ lowered the overpotential. This synthetic strategy of doping high-valent metal ions provides a new avenue for designing high-performance electrocatalysts.
Metal organic frameworks (MOFs), consisting of metal ions or clusters coordinated to organically linked molecules, have been of interest in the field of electrochemical research since the last few decades.25,26 Due to their structural flexibility and rich porosity, electrocatalysts formed by MOFs as templates usually present rapid diffusion of mass and fast transfer of electrons, therefore resulting in high catalytic activity.27,28 The most typical example is the conversion of MOF substrates to oxyhydroxides, which are also currently recognized as the true catalytically active species for most OER catalysts.29,30 For example, Tian et al. synthesized a CoFe-MOF and converted it into FeCo-oxyhydroxide with the assistance of electric fields, resulting in high OER activity.31 The element Hf can be oxidized to stable Hf4+, which is in a high valence state and has the potential to improve the electrocatalytic performance when doped into suitable substrates.32 However, the atomic radius, electronic structure, and coordination environment of Hf4+ are far from those of common transition metal atoms such as Co and Fe. It is somehow difficult to dope Hf4+ into hydroxides or oxyhydroxides containing Co and Fe while maintaining the catalyst structure and high accessibility. Therefore, it is crucial to find an excellent precursor to embed Hf for highly active electro-catalysis.
Therefore, in this work, we designed a unique etching procedure followed by electrosorption to fabricate a kind of trimetal oxyhydroxide/hydroxide composite (Fe,Hf)CoOOH/Co(OH)2 based on an easily obtainable Co-MOF. Briefly, the layer structure of the Co-MOF was gradually exfoliated in an alkaline electrolyte containing HfCl4 as a Lewis acid and recombined into a layered oxyhydroxide–hydroxide composite (Hf)CoOOH/Co(OH)2via electrochemical activation. Subsequently, Fe3+ was doped into the (Hf)CoOOH/Co(OH)2 composite by anodic electrochemical adsorption to form highly active (Fe,Hf)CoOOH/Co(OH)2. The synthetic procedure is shown in Scheme 1 and the resulting (Fe,Hf)CoOOH/Co(OH)2 exhibits distinguishing performance in OER catalysis.
To observe the details of the samples more clearly, we used ultrasonic equipment to peel the nanosheets from the carbon paper. The morphology and structure of the (Hf)CoOOH/Co(OH)2 and (Fe,Hf)CoOOH/Co(OH)2 samples were observed by transmission electron microscopy (TEM) and high-resolution transmission electron microscopy (HRTEM). As shown in Fig. 1c1 and c2, (Hf)CoOOH/Co(OH)2 and (Fe,Hf)CoOOH/Co(OH)2 exhibit normal nanoflake structures. The nanosheets are slightly broken because of ultrasonic exfoliation. The HRTEM images demonstrate the lattice properties of the two samples, as shown in Fig. 1d1, e1, d2 and e2. Both (Hf)CoOOH/Co(OH)2 and (Fe,Hf)CoOOH/Co(OH)2 exhibit two main lattice stripes which belong to the (101) planes of CoOOH (PDF: 1731), respectively. The corresponding SAED plots, shown in Fig. 1f1 and f2, exhibit two distinct arrays of CoOOH and Co(OH)2, proving that hydroxides and oxyhydroxides co-exist in the obtained catalysts. In addition, the SAED pattern of (Hf)CoOOH/Co(OH)2 exhibited good diffraction arrays, indicating that the introduction of Hf4+ did not affect the crystal structure of the oxyhydroxide and hydroxide. Furthermore, electrochemical adsorption of Fe3+ did not trigger significant crystallographic transformation, suggesting that most of the Fe3+ is incorporated into the catalyst matrix at the atomic scale. Fig. 1g1 and g2 show that O, Co, Fe and Hf are uniformly distributed in both catalyst matrices. This also illustrates that the electrochemical-Lewis co-acid etching method is effectively capable of stabilizing the anchoring of Hf4+ in the oxyhydroxide/hydroxide complexes.
In order to avoid the interference from chlorine gas evolution or Hf4+ reduction during the electrochemical-Lewis acid co-etching, the etching potential was set to 0–0.6 V (vs. Hg/HgO). The cyclic voltammetry (CV) result of the Co-MOF via electrochemical-Lewis acid co-etching in a mixed solution of KOH and HfCl4 is shown in Fig. 2a. Two recognizable redox pair peaks are observed at around 0.25 and 0.2 V, which can be attributed to the redox of Hf and Co, respectively. During the co-etching process, Co ions released from the Co-MOF combine with Hf4+ and OH− in the electrolyte and convert to hydroxide and subsequently to oxyhydroxide. Furthermore, chloride ions released from HfCl4 can form ion pairs with Co ions and stabilize the orthotetrahedral coordination structure of Co(OH)2, which facilitates the formation of α-Co(OH)2.33 Under cathodic current conditions, Hf4+ can diffuse toward the working electrode, adsorbing on the CoOOH/Co(OH)2 surface, integrating into the matrix or replacing part of the Co atoms, finally forming (Hf)CoOOH/Co(OH)2.
When the co-etching is completed, almost all Co-MOFs are converted to (Hf)CoOOH/Co(OH)2. However, excessive etching will destroy the coordination structure of (Hf)CoOOH/Co(OH)2, forming β-Co(OH)2 with lower electrochemical activity. In addition, excessive electrochemical Lewis acid co-etching leads to the disruption of the interwoven layered nanostructures and hinders the accessibility of the active sites. As shown in the inset of Fig. 2a, the sample subjected to deep etching also exhibits a color change. Therefore, based on the above analysis, we selected the sample after 20 cycles of etching as the precursor for the subsequent reactions. The precursors were washed with deionized water to remove Cl ions to prevent undesired chlorine gas evolution reactions at high potentials. Furthermore, as shown in Fig. 2b, there exists an optimal electrosorption state at 1150 s when doping Fe into the (Hf)CoOOH/Co(OH)2 catalyst. The maximum voltage limit of 4 V was not reached in the “optimal electrosorption state”; therefore, the voltage drop after this moment is triggered by oxygen evolution instead of discharge. After that moment, (Fe,Hf)CoOOH begins the oxygen evolution reaction, which is reflected by the gradual decrease in the potential required to maintain a constant current.
To investigate the effect of etching time on the conversion of the Co-MOF, we recorded SEM images at different cycle numbers. As shown in Fig. S2a,† it can be concluded that the Co-MOF slightly dissolved after the 5th cycle and the interwoven nanosheets gradually dissolved. After the 10th cycle, the structure of the Co-MOF further broke; notably, some Co-MOFs were exfoliated by etching and the layered structures began to emerge (Fig. S2b†). When the 20th cycle of CV was completed, Co-MOFs almost converted into layered nanostructures, with some lamellar materials arranged disorderly, intertwined and interpenetrated, but without forming the conventional nanoflower (Fig. S2c†). If the cycle exceeds 50 cycles, the excessively co-etched sample is almost dissolved and only a small amount of flake structure remains and the layered structure almost disappears drastically (Fig. S2d–f†).
In order to investigate the effect of doping Hf4+ on the crystal structure of oxyhydroxide/hydroxide composites, we prepared Hf-free samples CoOOH/Co(OH)2 and (Fe)CoOOH/Co(OH)2 according to the above electrochemical procedure and analyzed them by the X-ray diffraction (XRD) technique. At first, comparison of the XRD patterns of the four catalysts and Co-MOF (Fig. S3a†) reveals that the Co-MOF precursors are fully converted to the oxyhydroxide/hydroxide composites. In Fig. 2c, the XRD patterns of CoOOH/Co(OH)2, (Hf)CoOOH/Co(OH)2, (Fe)CoOOH/Co(OH)2, and (Fe,Hf)CoOOH/Co(OH)2 are compared. The sharp peaks at 25.1 and 43.5° indicate the carbon paper substrate. There are three peaks at 19.1, 33.4, and 38.1°, corresponding to the planes of Co(OH)2, and three peaks at 20.1, 36.9 and 38.9° correspond to the planes of CoOOH.
In particular, the XRD pattern of the carbon paper after the electrochemical process is almost the same as that of the bare carbon paper (Fig. S3b†), indicating that the carbon paper can be used as an effective substrate for electrocatalysts. In addition, a minor amount of other compounds is formed during the electrochemical conversion of the Co-MOF to oxyhydroxide/hydroxide. A small amount of FeOOH (JCPDS: 76-0182) was generated during the electrosorption, while (Hf)CoOOH/Co(OH)2 and (Hf,Fe)CoOOH/Co(OH)2 exhibit trace amounts of HfO2 (JCPDS: 81-0028). Also, a small amount of α-CoOOH (26-0480) was detected in all samples of catalyst. A comparison of the XRD patterns of the four catalysts and Co-MOF (Fig. S3†) reveals that the Co-MOF precursors are fully converted to the oxyhydroxide/hydroxide composites.
Fourier transform infrared (FT-IR) spectra of different catalysts were recorded to analyze the surface functional groups (Fig. S4†). The peaks at 1483 and 519 cm−1 are assigned to the metalO double bond in oxyhydroxide and metal–O stretching vibration, respectively. For the hydroxide, the peak at 635 cm−1 can be assigned to the metal–OH stretching vibration. In addition, the peaks located at 1600 and 1376 cm−1 are assigned to OH bending in H2O molecules and CO32−, respectively.34 Notably, the two samples of (Fe)CoOOH/Co(OH)2 and (Fe,Hf)CoOOH/Co(OH)2 that undergo electrosorption of Fe3+ exhibit strong metal–O peaks. This is due to the generation of defects and changes in layer spacing, which also promote the electrocatalytic activity. In addition, the metal–N peaks were absent in all four catalysts, which also indicates that the Co-MOF is fully converted into the target catalyst.
In order to investigate the effect of the introduction of Hf4+ and Fe3+ on the electronic states of the materials, we analyzed (Fe)CoOOH/Co(OH)2, (Hf)CoOOH/Co(OH)2 and (Fe,Hf)CoOOH/Co(OH)2 using X-ray photoelectron spectroscopy (XPS). Fig. S5† reveals the full XPS survey, clearly showing typical peaks at binding energies of 18, 198, 286, 531, 716 and 780 eV, corresponding to the elements Hf, Cl, C, O, Fe, and Co, respectively. The high-resolution XPS for Co 2p3/2 in (Fe)CoOOH/Co(OH)2 is shown in Fig. 2e. The peaks located at 783.0 and 781.2 eV represent Co2+ and Co3+ respectively. Due to the specific electro-etching process, the amount of Co3+ in both materials is much higher than that of Co2+. The high-resolution XPS spectra of Hf in (Hf)CoOOH/Co(OH)2 and (Fe,Hf)CoOOH/Co(OH)2 samples suggested the presence of Hf4+. As can be seen in Fig. 2f, two associated peaks of Hf are located at 16.84 and 18.51 eV, which are classified as 4f7/2 and 4f5/2, respectively.35 As shown in Fig. 2g, the high-resolution XPS spectra of O show three peaks attributed to metalO, metal–OH, and adsorbed oxygen in oxygen vacancies in all the three samples. In comparison with the other two samples, the apparent downward binding energy shift of the O atom is observed due to the introduction of Hf4+, and similarly, the binding energy shift of Co is also demonstrated in Fig. 2e. The XPS results demonstrate the electronically tuned formation of (Fe,Hf)CoOOH/Co(OH)2 with an optimal electronic structure for the OER electrocatalysis.36 The XPS high-resolution spectra of C demonstrate three chemical environments, C–C at 284.8 eV, C–O at 288.7 eV, and CO at 286.2 eV, in all the above three samples (Fig. 2h). The peak of the C–C bond may originate from the free ligand and the carbon paper substrate. Fig. S6† shows the XPS high-resolution spectra of Fe, which demonstrate the successful doping of the Fe element into (Fe)CoOOH/Co(OH)2 and (Fe,Hf)CoOOH/Co(OH)2 due to electrosorption.
The OER electrocatalytic activity of (Fe,Hf)CoOOH/Co(OH)2 catalysts was tested using a standard three-electrode cell system in 1.0 mol L−1 KOH solution at room temperature. In order to demonstrate the importance of Hf4+ in the catalysts, we comparatively tested CoOOH/Co(OH)2, (Hf)CoOOH/Co(OH)2, and (Fe)CoOOH/Co(OH)2. The LSV measurement used to test the OER activity is shown in Fig. 3a The overpotentials (η10) of the above four catalysts in descending order are (Fe)CoOOH/Co(OH)2 (203 mV) < (Fe,Hf)CoOOH/Co(OH)2 (250 mV) < CoOOH/Co(OH)2 (295 mV) < (Hf)CoOOH/Co(OH)2 (337 mV). As a result, the introduction of Fe3+ is able to greatly reduce the overpotential during the OER process. Even though the overpotential of (Fe,Hf)CoOOH/Co(OH)2 is higher than that of (Fe)CoOOH/Co(OH)2, (Fe,Hf)CoOOH/Co(OH)2 possesses higher activity at potentials beyond 1.51 V (marked as a pink dashed line), with an overpotential of 379 mV for a current density up to 100 mA cm−2. Similarly, the comparison of CoOOH/Co(OH)2 and (Hf)CoOOH/Co(OH)2 also shows the same result. Therefore, the Hf-doped catalysts show a delayed increase in current density, and Hf4+ is able to significantly improve the kinetics during the OER process. Fig. 3b shows that the Tafel slopes of all the samples and that of (Fe,Hf)CoOOH/Co(OH)2 are greatly reduced by the modulation of Hf4+. In addition, the Tafel slope plot at a high potential shows a similar result (Fig. S7†), suggesting that the Hf-doped material possesses excellent electrochemical activity. Notably, the OER catalytic activity of (Hf,Fe)CoOOH/Co(OH)2 also exceeds those of many previously reported Co or Fe-based electrocatalysts under similar conditions as presented in Table S1.†
In order to investigate the effect of the content of Hf4+ on the catalytic activity, we explored the OER activity of the catalysts with different doping amounts of Hf4+. (Hf)CoOOH/Co(OH)2-0.75, (Hf)CoOOH/Co(OH)2-1.5 (the abovementioned sample (CoHf)OOH/Co(OH)2), and (Hf)CoOOH/Co(OH)2-2.25 were obtained via etching under identical electrochemical cycles with different HfCl4 concentrations of 0.75 mmol L−1, 1.5 mmol L−1, and 2.25 mmol L−1, respectively. Hf4+ and Cl− modulate the coordination environment of CoOOH/Co(OH)2, and excessive Lewis acid (HfCl4) has a great influence on the conversion of the Co-MOF. As shown in Fig. S8,† the electrolyte with excessive HfCl4 dosage leads to almost complete dissolution of the Co-MOF, instead of being converted to hydroxide. Afterwards, we doped Fe into different Hf-doped samples to obtain the corresponding (Fe,Hf)CoOOH/Co(OH)2-x and examined their OER performance. As shown in Fig. S9,† the electrochemical properties of the obtained samples are different due to the various Hf4+ dosages in the samples. A higher content of Hf4+ doping (2.25 mmol L−1) would decrease the OER activity of the catalyst. Therefore, optimal Hf4+ dosage can significantly improve the OER catalytic activity.
The charge transfer impedance (Rct) can reflect OER kinetics. As can be seen in Fig. 3c from the Nyquist plots, they all possess low and similar impedance values, in which the charge transfer impedance of (Fe)CoOOH/Co(OH)2 and (Fe,Hf)CoOOH/Co(OH)2 is less than 0.15 Ω. This is because the embedding of ions in the Co(OH)2 interlayers improves the electrical conductivity and accelerates the charge transfer rates. Considering that the electrochemically active surface area (ECSA) is positively correlated with the double layer capacitance (Cdl), we calculated Cdl and ECSA for all four electrocatalysts by collecting the CV distributions in the non-Faraday region. Fig. 3d shows the CV profiles of all four catalysts at 12 mV s−1 sweep rate, and Fig. 3e illustrates the fitted Cdl values, which illustrates that the doping of Hf4+ and Fe3+ increased the Cdl values of the catalysts. All four catalysts exhibit high Cdl values, which suggests that the laminar structure provides sufficient active sites for the catalytic reaction. The CV curves of CoOOH/Co(OH)2, (Hf)CoOOH/Co(OH)2, (Fe)CoOOH/Co(OH)2, and (Fe,Hf)CoOOH/Co(OH)2, which were measured at different scan rates of 2–12 mV s−1 in the voltage range −0.7 to −0.6 V versus RHE, are displayed in Fig. S10.† Furthermore, the LSV and Tafel slope information normalized by ECSA is shown in Fig. S11,† which demonstrates better OER kinetics of (Fe,Hf)CoOOH/Co(OH)2. Long-term durability is another important parameter for evaluating OER catalysts; therefore, the activities of the catalysts were evaluated by performing continuous OERs in 1 mol L−1 KOH electrolyte at a constant current. As shown in Fig. 3f, all the four catalysts exhibited high stability, and no obvious potential fluctuations were found during the 15 h of continuous OERs, which indicates that the catalytic activity of oxyhydroxide/hydroxide composites is not significantly affected by phase transformation.
In order to further explore the effect of the doping of Hf4+ and Fe3+ on the catalytic activity of (Fe,Hf)CoOOH/Co(OH)2, density-functional theory (DFT) calculations were carried out based on the established crystal models. The model was built by geometry optimization of the 2 × 2 × 1 CoOOH supercell (JCPDS 14-0673). As shown in Fig. 4a, the charge density difference plots of (Fe,Hf)CoOOH indicate that the introduction of Hf4+ results in a redistribution of the local charge density. In this result, Fe and Co sites show local electron accumulation, while Hf exhibits local electron depletion. Fig. 4b and Fig. S12† illustrate the electronic states between (Fe,Hf)CoOOH/Co(OH)2 and (Fe)CoOOH/Co(OH)2 with O* intermediates. It is obvious that the transfer of electrons from Co atoms to O* takes place, which benefits the evolution of oxygen-containing species. In addition, the partial density of states (PDOS) of (Fe)CoOOH/Co(OH)2 and (Fe,Hf)CoOOH/Co(OH)2 are presented in Fig. 4c. The high overlap of the Co 3d and Fe 3d orbitals with the O 2p orbitals suggests that both (Fe)CoOOH/Co(OH)2 and (Fe,Hf)CoOOH/Co(OH)2 possess strong adsorption of O species. Indeed, the center of the d-band of the active sites decreased from −2.13 to −2.81 eV (Fig. S13†) significantly after the introduction of Hf4+. The incorporation of Hf4+ resulting in a downward shift of the metal d-band center could appropriately weaken the strong affinity of the catalyst for OOH*, effectively balancing the adsorption energies of the key intermediates and thus promoting fast OER kinetics of (Fe,Hf)CoOOH/Co(OH)2. Furthermore, as shown in Fig. S14,† the projected density of states of (Fe)CoOOH/Co(OH)2 and (Fe,Hf)CoOOH/Co(OH)2 demonstrate that the introduction of Hf4+ changed the electronic state. This change results in enhanced spin polarizability of (Fe,Hf)CoOOH/Co(OH)2 (Fig. S15†), which may enhance the catalytic activity.
When U = 1.23 V, as shown in Fig. 4d, the first step (*OH adsorption) of the OER is a spontaneous process at the Co and Fe sites in (Fe,Hf)CoOOH/Co(OH)2.37 The formation step of *OOH exhibits a high ΔG value, indicating that this is the rate-determining step of the OER. For the O2 release step, the ΔG of (Fe,Hf)CoOOH/Co(OH)2 is lower than that of (Fe)CoOOH/Co(OH)2 (Fig. S16†), suggesting that OOH * species are activated by Hf4+, verifying that the incorporating of Hf4+ indeed expedites the reaction kinetics of the OER.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4qi00823e |
This journal is © the Partner Organisations 2024 |