Bimetallic Doped RuO2 with Manganese and Iron as Electrocatalysts for Favorable Oxygen Evolution Reaction Performance.

Synthesizing oxygen evolution reaction (OER) catalysts with enhanced activity by codoping has been proven to be a feasible approach for the efficient use of noble metals via renewing their basic intrinsic properties. In continuation of the research in codoping, we prepare a ruthenium-based bimetallic doped catalyst Mn x Fe y Ru1-x-y O2 with an outstanding OER activity as compared to pure RuO2, one of the state-of-the-art OER catalysts. The synthesized codoped RuO2 with a Mn/Fe molar ratio of 1 reflected a Tafel slope of only 41 mV dec-1, which is appreciably lower than 64 mV dec-1 for pure RuO2. The X-ray photoelectron spectroscopy (XPS) performed reveals that oxygen vacancy and manganese valency are the key factors of the OER activity for codoped catalysts.

Introduction

Electrochemical water splitting has attracted global attention because it can transform renewable energy sources such as solar and wind energies into the high-power-density storable energy of hydrogen (H2), which is the major power source for fuel cells.[1−4] Polymer electrolyte membrane water electrolysis (PEMWE) is a promising technology for the production of H2, which offers the collection of hydrogen in a strong acidic environment at high pressure (over 150 bar) without additional compression.[5,6] Moreover, PEMWE maintains high efficiency at lower temperature, better stability, ionic conductivity, and pressure tolerance.[7,8] Commercial water electrolyzers require high functioning voltage, which is mainly associated with the sluggish kinetics in the half-cell anodic reactions. The oxygen evolution reaction (OER) activity of metal oxide catalysts is governed by the types and electronic structure of metal–oxygen bonds.[9,10]

However, in most of the cases, using noble metal oxide catalysts with simultaneous activity and stability is the only choice to meet the requirement of overpotential values in PEMWE.[11] As two representative noble metal oxides, RuO2 and IrO2 are most often chosen to be utilized as anodic catalysts resulting from their high catalytic activity in acidic media.[9,12,13] The high expense of their core elements, especially iridium, prevents themselves from large-scale industrial utilization with respect to impractical and uneconomical aspects.[14,15] Hence, we choose the more economical catalyst RuO2 with a reduced catalyst loading as a practical solution for business application.

Many efforts have been dedicated to the improvement of Ru-based electrocatalysts in the decrease in their content with a further enhancement in efficiency.[16,17] The doping of nonprecious elements into noble metal oxides has been proven feasible in the aspects of intrinsically increasing the OER activity and relatively minimizing the noble metal consumption.[18,19] The desired improvement in the activity is due to the change in the electronic density of the ruthenium element owing to the generated oxygen vacancies in the process of insertion of dopants into the ruthenium lattice.[10,20]

Therefore, in this study, the bimetallic doping is presented using two different 3d transition metal elements. The above-mentioned approach efficiently reduces the noble metal content and provides an excellent catalytic activity and stability. Until now, in most of the conducted research, transition metals have been extensively chosen as doping elements with Ru.[21,22] We choose manganese and iron as the doping transition metal elements due to their relatively better interactive performance and lower price, which are promising to achieve the prospective task. In addition, the concerned elements have been proven to help improve the electrocatalytic performance in the case of individual doping into the IrO2 lattice.[23,24] According to the element ratio, the composites are identified as MnFeRu-x%, where x is the Ru percentage (x = c(Ru)/[c(Mn) + c(Fe) + c(Ru)]). The electrochemical tests conducted for synthesized bimetallic doped composites exhibit higher OER catalytic activity and excellent stability in an acidic environment. The synergistic effect of dopants is the key factor for the enhancement of rutile phase catalysts in the OER activity. Thus, we discuss the guideline for such highly desired changes with respect to codoping perspectives of the foreign elements into the noble metals in the field of electrochemistry.

Experimental Section

Materials Synthesis

The different ratios between Ru, Mn, and Fe were achieved by varying the concentration of the precursor. The chemicals were of analytical grade and were used as received without further purification. Mn(NO3)2 and NaOH were supplied by Shanghai Lingfeng Chemical Reagent Co., Ltd. Fe(NO3)3 was supplied by Sinopharm Chemical Reagent Co., Ltd. RuCl3 was supplied by Shanghai Aladdin Bio-Chem Technology Co., Ltd. The autoclave model (50 mL) was supplied by Dongtai Zhongkaiya Stainless Steel Product Factory. In a typical hydrothermal procedure, Mn(NO3)2, Fe(NO3)3, and RuCl3 with required stoichiometric amount were mixed in 40 mL autoclave. Then 5 mL of 0.5 M NaOH and deionized water were added, resulting in the successive formation of indissoluble solid. The mixture was aged for 10 min without the addition of surfactant in this system to gain a greener synthesis approach. The autoclave was sealed tightly and slowly heated in an oven to 150 °C. The heating rate of the oven was 10 °C min–1. The solution was kept for 10 h to obtain fine crystals and then naturally cooled down to room temperature. The precipitate was collected, suction filtered, and washed several times with deionized water to remove residual ions. The retentate on the filter was dehydrated in an oven at 80 °C for 1 h. The final sample was collected and calcined at 600 °C for 6 h to produce excellent crystallinity. As for RuO2, the synthetic procedure is similar to the preparation of MnFeRu composites, with the only difference being no precursor of doping elements added.

Characterization

The crystallite structure of the prepared catalysts was investigated using D/max2550 V apparatus with a Cu Kα radiation source (λ = 1.5406 Å), and the generated data were recorded at a step size of 0.02° over a range of 10–80°. The field-emission scanning electron microscope (FESEM) was equipped with a Nova NanoS and energy-dispersive X-ray (EDX) spectrometer. The EDX spectrometer was used to observe the morphologies of the prepared catalysts and confirm the composition using a TEAMApollo system. Transmission electron microscopy (TEM) was undertaken with a JEOL Model 2100 LaB6 instrument, operating at 200 kV. Inductively coupled plasma (ICP) experiments were conducted through microwave digestion treatment. The surface properties of the catalysts were determined via X-ray photoelectron spectroscopy (XPS) using an ESCALAB 250Xi instrument with an Al Kα radiation source.

Electrochemistry

The OER performance of composite materials was investigated by using electrochemical techniques since oxygen evolution behavior is directly proportional to electron transference. The catalysts (weighted at 6.0 mg) were ultrasonically dispersed in 1.5 mL of isopropanol/deionized water (2:1 v/v) with 15 μL of 5% Nafion solution as a solvent. The Ti plates (0.5 cm × 1.5 cm) were used as support and precleaned by etching in 10 wt % faint boiling oxalic acid for 2 h. Then 7.5 μL of each catalyst ink was taken and drop-casted onto the precleaned Ti plate. The catalyst layer on the Ti plate was dried at 60 °C for 10 min. The process was repeated five times, leading to the total loading at 0.2 mg cm–2.

Electrocatalytic measurements were carried out with a standard three-electrode electrochemical cell controlled by a CHI660E workstation. A saturated calomel electrode (SCE) and a polished Pt foil were used as the reference electrode and the counter electrode, respectively. At the beginning of the electrocatalytic study, the electrodes were cycled between −0.2 and 1.4 V (vs SCE) at a scan rate of 50 mV s–1 until the curve values stabilized, leading to steady surface conditions. The polarization curves were obtained by extending the potential up-limit to 1.4 V (vs SCE) scanning at specific rates.

Results and Discussion

Structure and Morphology

The initially applied characterization techniques of X-ray diffraction (XRD) for the concerned composites help us explore the lattice structure along with other associated parameters. Figure shows the XRD patterns of all MnFeRu-x% composites (0.5 ≤ x ≤ 1), which account for the universality of the rutile structure with an obvious shift of peaks toward higher angles relative to the diffraction obtained from pure RuO2.

Figure 1

(a) X-ray diffractogram patterns of MnFeRu-x% composites. (b, c) Selected angle ranges showing the deviation extent for respective planes. (d) EDS spectra for MnFeRu-90. Atomic percent of all elements tested by ICP is shown in the spectra.

As Bragg’s law has confirmed, the diffraction angle is inversely proportional to the lattice constant, which further corresponds to the axial change in the crystal structure. According to our previous studies, the axial variation in the different stated planes brings benefit in enhancing the OER activity, while the rutile phase is well maintained.[9,18]

As shown in Figure , the XRD of Ru(MnFe)1–Oδ composites shows the rutile structure in their respective oxide constitution, which is consistent with the previous research.[9] In addition, the specific peak in Figure at 33.1° belongs to Fe2O3 for the planes (104), which corresponds to PDF chart no. 656920. Such dopants accommodate the limit of host structure when the substituted fraction proportion is below 30%. The maintenance of the solid solution in the rutile phase lattice system is actually governed by the elastic energy in the strained lattice of the host crystal structure.[25] Consequently, these observations clearly demonstrate that bimetallic doping can accommodate the Mn and Fe elements in the RuO2 lattice as a feasible approach.

The scanning electron microscopy (SEM) images of MnFeRu-90 are displayed in Figure with different resolutions, which give an insight into the lattice structure at a microlevel. The shape of the obtained catalyst confirms the formation of a prismatic-block-type rutile lattice. To further unveil the morphology of the mixed oxide composites, TEM was performed, as shown in Figure , which reveals the lattice fringes of the diffraction faces (110) in MnFeRu-90 and the corresponding d-spacing.[18]

Figure 2

(a, b) SEM images of MnFeRu-90. (c, d) Typical TEM images of MnFeRu-90. The insets in (d) showing fast Fourier transforms (FFTs) of the selected regions in the HRTEM, and the lattice fringe images corresponding to the selected regions.

To clearly trace the distribution of individual elements, FESEM and EDX elemental mapping images are illustrated in Figure . Red, green, yellow, and purple colors are respectively assigned to the oxygen, manganese, iron, and ruthenium elements. The FESEM patterns further confirm the uniform dispersion of the Mn and Fe elements with no existence of the Fe2O3 and Mn2O3 crystals, because the EDX elemental mapping shows rare iron- or manganese-rich regions, which is consistent with XRD results.

Figure 3

FESEM and EDX elemental mapping for MnFeRu-90.

Electrochemical Performance

The linear sweep voltammetry curves in Figure a exhibit the electrocatalytic activities of all prepared composites. The activity of the MnFeRu-90 sample is evidently higher compared with pure RuO2, as shown in the polarization curves for the OER process (Figure a). It can be clearly observed that the catalytic activity of a specific composite shows a considerable enhancement with the decrease of the noble metal content compared with pure RuO2, under the premise of almost no loss of stability. As illustrated in Figure f, the electrochemically active surface areas (ECSAs) for all composites are also tested by means of the CV measurements. The ECSA reveals the real electrode exposed area, and the changing trend of ECSA values shows that the key factor of the OER is an intrinsic activity.

Figure 4

(a) iR-corrected linear sweep voltammetry curves for the OER process of all codoped composites relative to RuO2 in the acidic solution (0.1 M HClO4). The horizontal dotted line marks the current density value of RuO2 at 1.7 V vs RHE. (b) Chronoamperometric curves at the constant potential of 1.7 V vs RHE. (c) The polarization curves for MnFeRu-90 before and after the durability test. (d) Tafel plots of RuO2 and MnFeRu-90. (e) Lines showing the mass activity of RuO2 and all codoped composites at 1.7 V vs RHE. (f) Electrochemically active surface areas (ECSAs) for all composites by extracting the double layer capacitance, Cd.

When it comes to the OER current response, the onset potential and Tafel slope are generally considered to be the key factors.[26,27] Therefore, as shown in Figure d, the OER activity difference between RuO2 and MnFeRu-90 composite can be annotated by the Tafel data of 65 mV dec–1 (RuO2) and 41 mV dec–1 (MnFeRu-90), with the onset potential maintaining the value of 180 mV, which is consistent with the earlier research.[9]

Given that the noble metal accounts for most of the cost, the ruthenium mass-specific activity makes sense. As shown in Figure e, when the Mn/Fe ratio (r(Mn/Fe)) maintains a value of 1 and the Ru proportion reaches the value of 90%, the mass–specific activity increases to the maximum for the Mn–Fe-codoped catalyst series. This result further illustrates that the electrocatalytic characteristics achieve the best performance with 90% Ru content for the codoped Ru-based catalysts in terms of cost and catalytic efficiency, which is in consistency with the previous study.[9]

Controlled voltage electrolysis was conducted so that the stability of RuO2 and the bimetallic doped catalyst (MnFeRu-90) could be determined. The prepared material coated on a Ti plate was kept in the 0.1 M HClO4 environment at a stable voltage of 1.58 V (vs RHE) for 5 h. As durability plots in Figure b,c show, the comparison of polarization curves before and after the stability test for pure RuO2 indicates a reduction of catalytic activity by around 20%, while the codoped catalysts show a similar durability performance as RuO2 during the course of the OER.[28] It is concluded from the above discussion that the specific codoped catalyst can withstand the corrosive environment along with its superb stability.

XPS Analysis

In line with the O 1s XPS for RuO2 and MnFeRu, Figures a and S5 show a shift toward lower binding energy (BE) for MnFeRu-90 and MnFeRu-80 with the higher BE for MnFeRu-50. The phenomenon demonstrates the presence of oxygen vacancies and a structural change in the surrounding ruthenium. The small decrease of O 1s BE values for MnFeRu-90 can be attributed to the participation of two codopants with different electronic environments. Especially, the shoulder peaks at around 531 eV are consistent with the reported oxygen vacancy peak for rutile phase anhydrous RuO2, which indicates a metallic character of MnFeRu.[29,30]

Figure 5

XPS spectra of (a) O 1s and (b) Ru 3d in RuO2 and MnFeRu-90. (c, d) Spectra of Fe 2p and Mn 2p in MnFeRu. The red peaks in (b) correspond to C 1s for calibration.

Electronic modulation in the rutile phase structure can also be derived from the varying Ru 3d BE for MnFeRu. As Figure b shows, the high-resolution spectra of Ru 3d for all ruthenium containing catalysts exhibit two peaks similar with those for pure RuO2 at 280.5 and 284.6 eV, corresponding to 3d5/2 and 3d3/2, respectively, which confirms the 4+ oxidation state of ruthenium along with the increase in specific peak values.[31] This may be owing to the attachment of RuO6 octahedron apical oxygen with the planar oxygen of each dopant octahedron on the augmentative scale. MnFeRu-90, with the 3d5/2 peak of relatively higher BE, exhibits the best catalytic character, which may be attributed to the difference of the extent of electronic modulation.

As shown in the high-resolution Mn 2p XPS (Figure d), the BE for all compounds are about 642.1 eV (Mn 2p3/2) and 653.9 eV (Mn 2p1/2), while the characteristic peak at 641.0 eV (Mn 2p3/2) disappears for MnFeRu-90. As previous studies suggest, octahedral MnO6 commonly exists as two types, Mn(1)O6 corresponding to the Mn4+ site and Mn(2)O6 related to the Mn3+ site, which can be probed through the relative position of Mn 2p3/2 and Mn 2p1/2.[29] Given that the BE interval value is 11.8 eV, the high valence Mn state is confirmed by the Mn 2p XPS spectrum. The low BE peaks in MnFeRu-80, MnFeRu-70, and MnFeRu-50 spectra correspond to Mn3+, while only one intensive de-convoluted peak is observed for MnFeRu-90. As the earlier studies have shown, the Ru ions may settle in the Mn(1) site due to the ionic radius and thus promote the OER efficiency for MnFeRu-90. As shown in the Fe 2p XPS (Figure c), main peaks at the BE values of 710.4 and 724.2 eV for Fe 2p3/2 and 2p1/2 can be identified in all composites, and the satellite peaks at about 718.3 eV can also be observed, which corresponds to previously reported +3 oxidation state of iron.[32,33] Though fluctuations in the BE of Fe 2p were observed with respect to their pure oxide forms, a trendless response was observed to have a direct correlation with the exhibited excellent OER performance of the catalysts.

The difference of the OER activity for bimetal-doped catalysts can also be explained via the study of the XPS result. According to O 1s spectra, the main feature for MnFeRu-90 is the peak representing the oxygen vacancy. The higher separate peak of MnFeRu-90 compared with pure RuO2 illustrates more exposure of oxygen defects at the catalyst’s surface. The higher valency of Mn for MnFeRu-90 observed from the Mn 2p spectrum also affects the arrangement of atoms and promotes the OER catalytic activity intrinsically.

Conclusions

Overall, our study confirms that manganese and iron can be feasibly utilized as two doping elements for rutile-type OER catalysts RuO2 in a 1:1 doping element molar ratio. MnFeRu-90 exhibits an enhanced catalytic activity compared with pure RuO2 while maintaining the stability. According to the results of XRD, SEM, and TEM, RuO2 maintains its crystal lattice structure when bimetallic doped with Mn and Fe dopants in a 1:1 ratio. The XPS data show that the higher oxidation state of the Mn site and the increase of oxygen vacancies bring about the modification of the host cationic metal. In this regard, codoping has a significant effect on the enhancement of the OER activity for the Mn/Fe-bimetal-doped RuO2 with doping elements with a molar ratio of 1. We believe that our conducted bimetallic doping approach of 1:1 molar ratio will bring benefit to the study of contributing factors in the OER to further reduce the noble metal content.