Novel MoSi₂ catalysts featuring surface activation as highly efficient cathode materials for long-life Li–O₂ batteries

Abstract

Searching for highly efficient and low cost electrocatalysts is an urgent challenge to provide high capacity and long cycle life Li–O₂ batteries (LOBs). In this work, the highly efficient electrocatalytic properties of MoSi₂ were first demonstrated using particles prepared by a solid–state reaction method. The surface was activated, playing a dominant role in the electrocatalytic performance. After acid etching to remove SiO₂ impurities arising from the surface oxidation, MoSi₂ exhibited excellent bifunctional electrocatalytic properties for LOBs. Mo⁶⁺ on the surface contributed to the initial low overpotentials of 0.45 V (3.2 V for the OER and 2.75 V for the ORR) and transformed into Mo⁴⁺ for a long cycle life. As a consequence, outstanding rate capabilities of 12 708 mA h g⁻¹ at 100 mA g⁻¹ and 10 794 mA h g⁻¹ at 800 mA g⁻¹, and excellent cycle durability of 215 and 90 cycles at a limited capacity of 1000 mA h g⁻¹ at 200 and 500 mA g⁻¹, respectively, were obtained for the MoSi₂ cathode.

---

Introduction

Li–O₂ batteries (LOBs), because of the potential large energy density of 3500 W h kg⁻¹, make themselves recognized candidates for the next generation of energy storage. Li–O₂ batteries mainly rely on the reversible oxygen evolution reaction (OER) and oxygen reduction reaction (ORR) on the cathode to achieve a charge/discharge process. Owing to the sluggish reaction kinetics, the incomplete decomposition of the discharge product Li₂O₂ and the formation of by-products lead to the increase of overpotentials, depressed round-trip efficiency, side reactions and the decomposition of electrolyte, and are attributed to the degradation of LOBs.^1,2 To tackle these problems, intensive efforts have been made to explore catalytic materials possessing outstanding electro–catalytic activity. Carbon materials with a superior electric conductivity and designed functional structure were firstly used to catalyze the electrode reaction and achieved an innovative breakthrough in improving the performance of LOBs.^3,4 However, carbon materials are unstable and oxidize in the discharging process. A side reaction occurs above 3.5 V during charging and forms Li₂CO₃, which causes a reduction in active substances and the production of by-products, reducing the reversibility and cyclic performance of the Li–O₂ battery. Moreover, the carbon materials promote the decomposition of the electrolyte (especially the widely used dimethyl sulfoxide (DMSO) and tetraethylene glycol dimethyl ether (TEGDME) electrolytes) during the charging process and produces by-products resulting in an increase of the overpotential decay of the capacity and reduction of the cyclic performance.^5,6 It's essential to develop an alternative catalyst for the Li–O₂ battery which could improve the OER or ORR process. Many efforts have been made to explore catalytic materials for LOBs, and only limited materials, such as precious metals, metal alloys, metal oxides, and porous metal carbides, exhibit outstanding catalytic capabilities.^7–10 Searching for outstanding catalytic materials with low overpotentials and stable cycle durability is still the most important challenge for LOBs.

Up to now, no works have reported the catalytic capability of silicide for LOBs, even though they possess a relative narrow band gap, good electric conductivity and high catalytic performance.^11,12 In the current contribution, we selected molybdenum disilicide (MoSi₂) as the cathode material and demonstrated the outstanding electrocatalytic capability of MoSi₂ for LOBs. MoSi₂ is a Dalton-type intermetallic compound, and the atomic combination in the crystal structure exhibits the characteristics of the coexistence of metal bonds and covalent bonds,^13,14 and has excellent high-temperature essential characteristics. Due to its excellent oxidation and corrosion resistance, MoSi₂ has been applied as a structural material, electric heating element and protection coating.^15–17 Its high electric conductivity, narrow band gap and chemical stability are thought to be suitable as the catalytic materials for LOBs. Meanwhile, the surface oxidation of MoSi₂ (MoSi₂ + O₂ → MoO₃ + SiO₂) restricts its electrocatalytic performance in LOBs, in which the SiO₂ impurity is inactive for the ORR and OER processes in LOBs. Furthermore, high purity MoSi₂ powder is difficult to be prepare as the SiO₂ almost exists as an impurity in the synthesized MoSi₂ particles.

In this work, high purity MoSi₂ is prepared from a solid–state reaction method in a large scale following an acid etching process to delete the impurities. The catalytic characteristics of MoSi₂ as a cathode material for LOBs were investigated.

MoSi₂, as a bifunctional catalyst, can effectively catalyze both ORR and OER and remain stable during the electrochemical process. The discharge/charge evolution during cycling protected the surface oxidation of MoSi₂ for an excellent cycling durability. The low overpotentials in the charge/discharge process is found to be dominated by the surface condition of the Mo ion. As a consequence, the MoSi₂ electrode catalyst exhibited a remarkable electrochemical performance in terms of the low overpotentials of 0.45 V (cut-off voltage of 1000 mA h g⁻¹ at 200 mA g⁻¹), outstanding rate capability (12 708 mA h g⁻¹ at 100 mA g⁻¹ and 10 794 mA h g⁻¹ at 800 mA g⁻¹), and excellent cycle durability (215 cycles at the limited capacity of 1000 mA h g⁻¹ with a current of 200 mA g⁻¹).

Experimental section

Material preparation

In the typical solid reaction process, 18 g of MoO₂ (99% purity, Aladdin Chemistry Co. Ltd.) and 12 g Si of powder (99.9% purity, Sinopharm Chemical Reagent Co. Ltd.) were weighed and put into a mill pot with 1 ml of ethyl alcohol and ZrO₂ balls. Of these, ethyl alcohol was used as the solvent while the ZrO₂ balls were used as the grinding medium. The ratio of balls and powder was 20 : 1 and the rotating speed was 200 rpm. Ar was filled into the grinding jar as the protecting atmosphere. After mixing and grinding for 8 hours, the mixture was put in the oven for 8 hours at a temperature of 80 °C to remove the solvent. After drying, the powder was filtrated through a 60 mesh sieve to get a fine material. Then, the material was sintered at a heating rate of 10 °C min from room temperature to 1400 °C and kept at 1400 °C for 60 minutes. All the sintering process was conducted under an argon atmosphere. After sintering, the as-prepared MoSi₂ was obtained. In order to get purer MoSi₂, the powder was etched by hydrofluoric acid for 6 hours to remove the by-product SiO₂. And then, the etched MoSi₂ was cleaned by deionized water twice and then put in the oven at 80 °C for 12 h to remove the water.

Material characterization

The morphology of the material was characterized by a field scanning electron microscope (JEOL JSM-7500F) with an X-ray energy dispersive spectrometer (EDS). To observe the shape and microstructure of the material more clearly, TEM and HRTEM images were obtained by a transmission electron microscope (JEOL JEM-2011). The X-ray diffraction pattern (XRD) was tested to determine the phase composition by means of an X-ray diffractometer (Rigaku D/Max-r B) with Cu-Kα (λ = 1.540 Å) radiation with a voltage of 40 kV and a current of 40 mA. Raman spectra were obtained to determine the composition and molecular structure by a Bruker spectrometer with a laser of 532 nm. X-ray photoelectron spectroscopy (XPS, ESCALAB 250XI) was also used to determine the chemical composition of the sample.

Electrochemical measurement

First, the as-prepared catalyst accompanied with Ketjenblack and poly(tetrafluoroethylene) were mixed at a mass ratio of 4 : 4 : 2, and then 3 ml of isopropanol was added as the solvent. Ultrasonic techniques were used to spread the solution evenly. Then, the solution was distributed into carbon paper evenly. Each time, 10 mg of catalyst, 10 mg of Ketjenblack and 5 mg of poly(tetrafluoroethylene) were weighed, and spread over 10 cathodes uniformly. The average mass of the catalyst in each cathode was about 1 mg. To remove isopropanol, the carbon paper was put into a vacuum oven for 12 hours. The difference in weight of the carbon paper before and after adding the solution was recorded to calculate the mass of the active catalyst. All the batteries were assembled in a glove box which was occupied with Ar (H₂O < 0.1 ppm, O₂ < 0.1 ppm). An integrated cell contained the Li anode, a glassy fiber separator, electrolyte, and catalyst attached to the cathode. In this work, the electrolyte was 1 M lithium bis(trifluoromethanesulfonyl) in triethylene glycol dimethyl ether (LiTFSI/TEGDME). The electrochemical performance was studied using an electrochemical workstation (PARSTAT MC). All the batteries were sealed in an oxygen filled bottle during each electrochemical measurement.

Result and discussion

The MoSi₂ powder was synthesized through a solid–state reaction method. Si and MoO₂ powder were firstly ground for 8 h, and then sintered under an Ar atmosphere for 1 h. To eliminate the SiO₂ impurity, the as prepared MoSi₂ powder was etched in 1 M HF solution for 6 hours. X-ray diffraction (XRD) was used to investigate the crystalline structure of the MoSi₂ particles. Fig. 1a shows the XRD pattern of MoSi₂ particles before and after acid etching. The diffraction peaks at 22.65, 30.10, 39.74, 44.68, 75.59° matched well to the (002), (101), (110), (103) and (213) planes of MoSi₂ (PDF#41-0612). For the MoSi₂ particles before acid etching, a diffraction peak of SiO₂ at 21.98° (PDF#39-1425) was detected, which indicated the existence of an SiO₂ impurity. The SiO₂ impurity will degrade the catalytic performance of MoSi₂.

Fig. 1 (a) The XRD patterns of MoSi₂ particles before and after acid etching. (b) A SEM image, (c) a TEM image and (d) a SAED pattern of MoSi₂. (e) Element content data, and EDS mapping of Mo, Si and O elements. (f) The body-centered tetragonal unit cell of MoSi₂.

A morphological observation was also carried out to identify the shape of MoSi₂ particles. As shown in Fig. 1b and c, the MoSi₂ particles showed an irregular morphology with a particle size that ranged from 500 nm to 2 μm. The diffraction spots in the selected area electron diffraction (SAED) patterns were identified as the (101), (110), (103), (200), (116) and (301) planes of the MoSi₂ particles, consistent with the XRD pattern. The EDS analysis shown in Fig. 1e indicated that the Mo/Si molar ratio was about 1 : 2 in the acid etched MoSi₂ particles, meanwhile, the Mo/Si molar ratio was only about 1 : 3 for the MoSi₂ particles before acid etching, which indicated the existence of SiO₂ impurities. The element mapping showed that Mo and Si elements were uniformly distributed in the MoSi₂ particles and almost no signal of the O element was detected. The element mapping of MoSi₂ without acid etching was also investigated (Fig. S1†). As shown in Fig. 1f, MoSi₂ is determined as a body-centered tetragonal phase^18,19 (I4/mmm, a = 3.2 Å, b = 3.2 Å, c = 7.8 Å). The bonds between the Si atoms are covalent bonds, Mo–Mo bonds are metal bonds, while Mo–Si bonds are between covalent bonds and metal bonds. Thus, the crystal structure has the common characteristics of metal bonds and covalent bonds, which subsequently makes MoSi₂ have the dual characteristics of ceramics and metals.

To further investigate the crystalline structure and the surface elemental composition, Raman spectra and X-ray photoelectron spectroscopic (XPS) measurements were performed. Raman spectra are shown in Fig. S2.† The vibration peaks located at 324 and 434 cm⁻¹ correspond to the A_1g and E_g modes of MoSi₂ and no other vibration peaks were detected, which indicates the formation of the pure phase of tetragonal MoSi₂.²⁰ Surface composition and the valence state of the MoSi₂ particles were identified by XPS analysis and are shown in Fig. 2a–c. The presence of the Mo, Si and O elements is confirmed in Fig. S3.† In the high resolution XPS spectrum of Mo 3d (Fig. 2a), the Mo 2d_3/2 and Mo 2d_5/2 peaks are located at binding energies of 231 and 227 eV, respectively, representing Mo⁴⁺ in the MoSi₂ lattice. The binding energy peak of 234.5 eV represents the existence of Mo⁶⁺.^21–23 In the high resolution XPS spectrum of Si 2p (Fig. 2b), the banding energy peak at 97.5 eV can be attributed to the Mo–Si bond while the other peak at 100 eV is associated with the Si–Si bond, which suggests that SiO₂ impurities were totally removed after acid etching.^24,25 In the high resolution XPS spectrum of O 1s (Fig. 2c), the fitting peak at 529 eV is assigned to the Mo–O bond, which confirmed the existence of MoO₃, and the peak at 531.9 eV is attributed to the existence of physically and chemically bonded water on the surface.^26,27 The pore size distribution and specific surface area were measured by BET N₂ adsorption/desorption. The N₂ adsorption/desorption curves show the typical features of a type IV isotherm with a H3-type hysteresis loop,^28–31 and the specific surface area was calculated to be 1.771 m² g⁻¹. From Fig. 2d, the pore size of the MoSi₂ material was concentrated at 2.8 nm. The presence of mesopores is thought to be caused by acid etching, and is beneficial for the transition of oxygen and Li⁺, but also provides numerous active sites for the formation of discharge products. The pore size distribution and N₂ adsorption/desorption curves of MoSi₂ before acid etching are shown in Fig. S4.† The surface area of the un-etched MoSi₂ was 1.916 m² g⁻¹, which is similar to that of acid etched MoSi₂. On the other hand, the pore size concentrated at 4.6 and 9.8 nm, respectively. After acid etching, the pore size distribution of MoSi₂ moved to a smaller size.

Fig. 2 High-resolution XPS spectra of etched MoSi₂ particles: (a) Mo 3d, (b) Si 2p, and (c) O 1s. (d) The pore size distribution and N₂ adsorption/desorption isotherms of the MoSi₂ particles.

Fig. 3a shows the CV curves of MoSi₂ particles before and after acid etching as electrodes between cut-off voltages of 2.35 V to 4.35 V for Li–O₂ batteries. For the MoSi₂ particles after acid etching, a reduction peak at 2.63 V and one oxidation peak at 3.44 V were observed, which brought out a stable electrochemical reaction platform. For the MoSi₂ particles before acid etching, two reduction peaks at 3.4 V and 4.0 V were observed. The MoSi₂ electrode exhibited larger cathodic and anodic current values, which indicated that the MoSi₂ particles displayed a better catalytic capability in electrode reactions, especially the OER process, after acid etching.³² EIS was also used to measure the catalytic performance of two samples. The charge transfer resistance was evaluated by the diameter of the middle frequency depressed semicircle.³³ As shown in Fig. 3b, the charge transfer resistance reduced from 30 Ω to 20 Ω after acid etching. It can be concluded that the conductivity of the MoSi₂ particles increased and more MoSi₂ active sites were exposed after removing the insulating SiO₂ impurities. The EIS of batteries which use MoSi₂ as the cathode catalyst were also tested after the discharge and recharge process. The charge-transfer resistance of the two batteries increased significantly after discharging. After recharging, the diameter of the EIS curve of acid etched MoSi₂ decreased to 37 Ω, while that of un-etched MoSi₂ remained 75 Ω. It can be concluded that the MoSi₂ particles conducted a better reversibility after acid etching.

Fig. 3 (a) CV curves and (b) EIS spectra of acid etched MoSi₂ and un-etched MoSi₂. (c) Initial discharge/charge profiles of acid etched MoSi₂, un-etched MoSi₂ and KB at 100 mA g⁻¹ over a voltage range of 2.35 to 4.35 V. (d) Initial discharge/charge plots of acid etched MoSi₂, un-etched MoSi₂ and KB with a fixed specific capacity of 1000 mA h g⁻¹ at 200 mA g⁻¹. (e) The cycling performance of acid etched MoSi₂ for selected cycles at 200 mA g⁻¹ with a fixed capacity of 1000 mA h g⁻¹. (f) Discharge/charge profiles after etching the MoSi₂ cathode at different current densities over the voltage range of 2.35 to 4.5 V. Cycling performance of MoSi₂ before etching, after etching and KB at (g) a fixed specific capacity of 1000 mA h g⁻¹ at 200 mA g⁻¹ and (h) a fixed specific capacity of 1000 mA h g⁻¹ at 500 mA g⁻¹. (i) A photograph of a working light emitting diode powered by the prepared LOBs using MoSi₂ particles as a catalyst.

As shown in Fig. 3c, the first discharge capacity of the MoSi₂ cathode after acid etching delivered extra-high specific capacities of 12 708 mA h g⁻¹ with a cut-off voltage window of 2.35 to 4.35 V at 100 mA g⁻¹, calculated based on the mass of MoSi₂ on the electrode, which was obviously higher than that of the particles before acid etching and of KB, 9186 and 4136 mA h g⁻¹, respectively. The MoSi₂ particles after acid etching showed an enhanced catalytic activity towards OER in LOBs with two charge platforms of 3.5 V and 4.2 V, respectively, at the first galvanostatic discharge/charge cycle, compared with that of KB (overpotential of 1.78 V). These results indicated that the impurity of SiO₂ in MoSi₂ particles degraded its catalytic performance for LOBs. The first charge/discharge curves, at a fixed capacity of 1000 mA h g⁻¹ and a 200 mA g⁻¹ current, of the MoSi₂ particles are also shown in Fig. 3d. The MoSi₂ particles after acid etching showed an enhanced activity towards ORR and OER in the LOBs since the charge/discharge potential was only 3.2 V and 2.75 V, respectively, with overpotentials of 0.24 V and 0.21 V at the first charge/discharge cycle, compared with that of KB (overpotentials of 2.72 V) and un-etched MoSi₂ electrode (overpotentials of 1.42 V).

The cyclic stability was examined at a current density of 200 mA g⁻¹ with a fixed capacity of 1000 mA h g⁻¹ and is shown in Fig. 3e. It is worth noting that the charge potentials changed largely in the first 5 cycles, in which the charge potentials increased from 3.2 to 3.4 V, and then increased slowly in the following cycle testing. The MoSi₂ electrode obtained an excellent cycle stability and maintained the capacity for almost 215 cycles. The terminal discharge/charge voltage (Fig. 3g) remained at 2.56 V and 4.29 V, respectively, which showed an excellent cycle durability and low over potentials after 215 cycles. This result is connected with the surface condition of MoSi₂, which is introduced in the following parts. In the same testing condition, the discharge platform of un-etched MoSi₂ electrode decreased to just 0.7 V after 24 cycles and that of the KB electrode decreased to 1 V after 14 cycles. When the MoSi₂ electrode was measured at a high current density of 500 mA g⁻¹ with a fixed capacity of 1000 mA h g⁻¹, it can still maintain the capacity for 90 cycles with discharge/charge potentials of 2 V and 4.9 V (Fig. 3h). Under a fixed capacity of 600 mA h g⁻¹ at a current of 100 mA g⁻¹, the MoSi₂ electrode could maintain the capacity for almost 270 cycles (Fig. S5†). Even after 270 cycles, the overpotential still maintained 2.05 V with a charge voltage of 4.31 V and discharge voltage of 2.26 V. Whereas, the MoSi₂ electrode before acid etching could only maintain the discharge voltage above 2.0 V for less than 140 cycles, and for KB, the voltage dropped to 1.5 V at the 15th cycle, suggesting that the acid etched MoSi₂ electrode possesses a much better stability and reversibility. Fig. 3i displays a photograph which records a working light-emitting diode powered by a Li–O₂ battery utilizing MoSi₂ as the electrode. Typical discharge/charge profiles of un-etched MoSi₂ and KB electrode under different current densities and cut-off specific capacities is displayed in Fig. S7 and S8,† respectively.

The rate capability was tested for the MoSi₂ particles after acid etching with a cut-off voltage window of 2.35-4.5 V at different currents, as shown in Fig. 3f. The discharge capacity decrease accompanied by the charge/discharge polarization at a high current density can be ascribed to the increase of the internal charge transfer resistance of the LOBs. Meanwhile, impressive specific capacities of 12 048 mA h g⁻¹, 11 336 mA h g⁻¹, and 10 794 mA h g⁻¹ can be still obtained by the MoSi₂ electrode at high current densities of 200 mA g⁻¹, 500 mA g⁻¹ and 800 mA g⁻¹, respectively, while the results MoSi₂ of without acid etching were not close to those of the MoSi₂ electrode, as recorded in Fig. S7d,† indicating the significant advantage of the high rate capability of MoSi₂ materials.

The charge process of the MoSi₂ electrode was further investigated by in situ differential electrochemical mass spectrometry (DEMS) to reveal the evolution of gas species. The cell was performed under a current density of 500 mA g⁻¹ with a fixed capacity of 500 mA h g⁻¹. The gas analysis curves of the O₂, H₂O and CO₂ evolution rates for the charging process are presented along with the galvanostatic charge voltage profile in Fig. 4. The plot of the O₂ evolution rate can be divided into three stages and presented a typical M-shape.^34–36 In Stage I, the O₂ gas was quickly released, and then decreased with the increase of charge voltage at stage II. This process was attributed to the fast decomposition of defective the Li_2−xO₂ phase at a low charge potential.³⁷ When the charge voltage reached approximately 4.2 V, the O₂ evolution rate gradually rose again accompanied with a stable charge voltage platform. This result can be attributed to the crystalline Li₂O₂ clusters or particles which decomposed at a higher voltage during charging. The DEMS results indicated that the discharge product was a mixture of nonstoichiometric amorphous Li_2−xO₂ and crystalline Li₂O₂ phases. It is worth noting that there is no release of H₂O during charge process, indicating a brilliant reversibility and cycle performance of the MoSi₂ electrode. A small amount of CO₂ was observed at the end of charging. This unexpected product may be caused by the inevitable side reaction between carbon and electrochemically unstable electrolyte.^38,39

Fig. 4 In situ DEMS curves of the MoSi₂ catalyst, which were collected during the charging process at a fixed specific capacity of 500 mA h g⁻¹ at a current of 500 mA g⁻¹.

Aiming at exploring the discharge and charge process clearly, the morphology and components of the discharged cathode were also tested. Fig. 5a is the SEM image of the fresh cathode, and only some particles can be observed. But after full discharge, a thick Li₂O₂ film covers the surface of the particles (Fig. 5b). The initial discharge product was also monitored when discharged at 100 mA h g⁻¹ and shown in Fig. S9.† A sheet-like discharge product was observed on the surface of the MoSi₂ particles. The formation of a sheet-like discharge product was beneficial to the mass transition and indicated a solvation-mediated growth mechanism of Li₂O₂ in the initial discharge stage.^40,41 O₂ underwent a one-electron reduction to O²⁻ on the surface of the catalyst, and then dissolved into the electrolyte. O²⁻ combined with Li⁺ in the solution to form the intermediate LiO₂ which was partially adsorbed to the surface of the electrode (LiO_2sur).⁴²Scheme 1 illustrates the mechanism for the electrocatalytic reactions during the discharge progress. After recharge, the film disappeared and the morphology was similar to that of the fresh cathode (Fig. 5c). Fig. S10† shows the SEM image of the MoSi₂ electrode after 40 cycles. No morphology change or non-decomposed discharge product was observed on the surface of the MoSi₂ particles, indicating the smooth decomposition of Li₂O₂ during cycling. The crystalline nature of the discharge products was analyzed using XRD measurements, and Fig. 5d displays the XRD pattern of the three cathodes. It can be seen that the peaks of the fresh cathode are similar to those of the recharged cathode. Compared to them, the discharged cathode has two peaks which are assigned to the (100) and (101) planes of Li₂O₂, and the results are totally consistent with the SEM test. Furthermore, the XPS test was also conducted to analyze the discharge product. Fig. S11a† manifests that the pattern of Li 1s can be divided into only one peak located at 55.4 eV, and it represent the existence of Li₂O₂. Besides, the EIS was measured to analyze the electronic transmission capability of the Li–O₂ battery. From Fig. S11b,† the electrochemical impedance is quite low, but after a CV test, the impedance improved greatly. This implies that the electrochemical reaction can produce an insulating product which impedes the transition of the electron inside the battery.

Fig. 5 SEM images of (a) a fresh cathode, (b) a discharged cathode, and (c) a recharged cathode. (d) XRD patterns of the MoSi₂ cathode: fresh, discharged, and recharged.

Scheme 1 A schematic illustration of the proposed reaction mechanism during the discharge process.

To investigate the reaction mechanism, ex situ XPS was further conducted to explore the detailed surface states of the MoSi₂ electrode during a full-charge/discharge progress at a current density of 100 mA g⁻¹, as presented in Fig. 6b–e, in accordance with states 1–4 in Fig. 6a. The Mo 3d XPS spectra can be deconvoluted into three peaks which indicated that the two major peaks of 3d_3/2 (231 eV) and 3d_5/2 (227 eV) are associated with Mo⁴⁺ while the peak at the high energy region (234.5 eV) can be attributed to Mo⁶⁺.^43–45 Compared to the pristine MoSi₂ particles (Fig. 2c), the intensity of Mo⁶⁺ decreased largely at the first state (Fig. 6b). Due to the formation of a thick Li₂O₂ layer, only weak peaks corresponding to Mo⁶⁺ and Mo⁴⁺ were detected in state 2 (Fig. 6c). In contrast, after recharging to the cut-off voltage of 3.6 V (state 3), the peak intensity of Mo⁴⁺ significantly increased, and the XPS of Mo 3d showed approximately the same Mo⁶⁺/Mo⁴⁺ intensity ratio compared with that in the fresh cathode at the end of fully charge (state 4).

Fig. 6 (a) The selected states in a voltage-limited discharge and charge process. High-resolution XPS spectra of Mo 3d at (b) state 1, (c) state 2, (d) state 3 and (e) state 4, corresponding to states (1–4) in (a), respectively. (f) The high-resolution XPS spectrum of the MoSi₂ cathode for Mo 3d after 10 cycles.

The elevating concentration of Mo⁴⁺ during charging was resulted from the constant decomposition of the discharge product on the cathode surface and thus more Mo⁴⁺ active sites were released. The high resolution XPS of Mo 3d after 10 cycles (with a fixed specific capacity of 1000 mA h g⁻¹ at a current of 200 mA g⁻¹), which corresponds to the selected state in Fig. S12,† was also investigated (Fig. 6f). It was apparent that the concentration of Mo⁶⁺ decreased compared to the initial particles even in the cycle testing.

It has been reported that Mo⁶⁺ has good OER catalytic capabilities and could significantly decreased the overpotentials in the ORR/OER process in LOBs.⁴⁶ These results indicated that Mo⁴⁺ was the main valence state for MoSi₂ during the charge/discharge process of LOBs. The consumption of Mo⁶⁺ in the initial discharge/charge cycle may be the cause of the gradual increase in charge potential during the start of several cycles, as introduced in Fig. 3e. The high resolution XPS spectra of Si 2p was also investigated (Fig. S13†); the peaks show almost no change, which indicated the outstanding stability of MoSi₂ catalyst.

Conclusions

In summary, MoSi₂ exhibited superior electrocatalytic performance as a cathode material for LOBs with acid etching removing the SiO₂ impurities. The acid etched MoSi₂ electrode obtained had a large discharge capacity, high rate capability and excellent cycle performance with low overpotentials. It can be concluded that the SiO₂ impurities are harmful to the improvement of the electrocatalytic performance of MoSi₂. The surface condition of MoSi₂ had a large influence on its electrocatalytic performance. A high concentration of Mo⁶⁺ decreased the overpotentials of MoSi₂; meanwhile, Mo⁴⁺ was the main valence state of MoSi₂ during cycling. In this work, we firstly demonstrated that silicide-based materials can demonstrate excellent electrocatalytic capabilities for LOBs and provide a practical way to find novel high performance cathode materials for LOBs.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Key R&D Program of China of 2017YFE0195200, Open Program of Tsinghua University State Key Laboratory of New Ceramic and Fine Processing (KF201814, KF201805), Open Program of Guangxi Key Laboratory of Information Materials (171002-K) and the China Scholarship Council.

Notes and references

1. L. Ma, T. Yu, E. Tzoganakis, K. Amine, T. Wu, Z. Chen and J. Lu, Adv. Energy Mater., 2018, 8, 1800348.
2. Y. Hou, J. Wang, J. Liu, C. Hou, Z. Xiu, Y. Fan, L. Zhao, Y. Zhai, H. Li, J. Zeng, X. Gao, S. Zhou, D. Li, Y. Li, F. Dang, K. Liang, P. Chen, C. Li, D. Zhao and B. Kong, Adv. Energy Mater., 2019, 1901751,  DOI:10.1002/aenm.201901751.
3. S. D. Beattie, D. M. Manolescu and S. L. Blair, J. Electrochem. Soc., 2009, 156, A44.
4. B. M. Gallant, R. R. Mitchell, D. G. Kwabi, J. Zhou, L. Zuin, C. V. Thompson and Y. Shao-Horn, J. Phys. Chem. C, 2012, 116, 20800–20805.
5. M. M. Ottakam Thotiyl, S. A. Freunberger, Z. Peng and P. G. Bruce, J. Am. Chem. Soc., 2013, 135, 494–500.
6. M. Balaish, J.-W. Jung, I.-D. Kim and Y. Ein-Eli, Adv. Funct. Mater., 2019, 1808303,  DOI:10.1002/adfm.201808303.
7. S. Ma, Y. Wu, J. Wang, Y. Zhang, Y. Zhang, X. Yan, Y. Wei, P. Liu, J. Wang, K. Jiang, S. Fan, Y. Xu and Z. Peng, Nano Lett., 2015, 15, 8084–8090.
8. J. Huang, B. Zhang, Y. Y. Xie, W. W. K. Lye, Z.-L. Xu, S. Abouali, M. Akbari Garakani, J.-Q. Huang, T.-Y. Zhang, B. Huang and J.-K. Kim, Carbon, 2016, 100, 329–336.
9. Q. Liu, Y. Jiang, J. Xu, D. Xu, Z. Chang, Y. Yin, W. Liu and X. Zhang, Nano Res., 2014, 8, 576–583.
10. G. Sun, Q. Zhao, T. Wu, W. Lu, M. Bao, L. Sun, H. Xie and J. Liu, ACS Appl. Mater. Interfaces, 2018, 10, 6327–6335.
11. M. Wu, Y.-n. Lin, H. Guo, T. Ma and A. Hagfeldt, J. Power Sources, 2014, 263, 154–157.
12. M. Seibt, R. Khalil, V. Kveder and W. Schröter, Appl. Phys. A: Mater. Sci. Process., 2008, 96, 235–253.
13. J. J. Petrovic, Mater. Sci. Eng. A, 1995, 192, 31–37.
14. R. S. Rastogi, V. D. Vankar and K. L. Chopra, J. Vac. Sci. Technol., A, 1992, 10, 2822–2825.
15. T. Murakami, H. Mano, Y. Hibi, K. Matsuzaki and H. Inui, Tribol. Int., 2014, 69, 61–69.
16. Y. Zhang, Y. Li and C. Bai, Ceram. Int., 2017, 43, 6250–6256.
17. J. Xu, Z. Li, Z.-H. Xie, P. Munroe, X. L. Lu and X. F. Lan, Appl. Surf. Sci., 2013, 270, 418–427.
18. O. Andersen, O. Jepsen, V. N. Antonov, V. Antonov, B. Y. Yavorsky, A. Y. Perlov and A. Shpak, Phys. B, 1995, 204, 65–82.
19. L. F. Mattheiss, Phys. Rev. B: Condens. Matter Mater. Phys., 1992, 45, 3252–3259.
20. D. Sciti, S. Guicciardi, G. Celotti, M. Deluca and G. Pezzotti, Adv. Eng. Mater., 2007, 9, 393–399.
21. C. Hou, J. Wang, W. Du, J. Wang, Y. Du, C. Liu, J. Zhang, H. Hou, F. Dang, L. Zhao and Z. Guo, J. Mater. Chem. A, 2019, 7, 13460–13472.
22. S. Zhang, G. Wang, J. Jin, L. Zhang, Z. Wen and J. Yang, Nano Energy, 2017, 36, 186–196.
23. L. Shaw and R. Abbaschian, J. Mater. Sci., 1995, 30, 5272–5280.
24. Y. Fu, H. Du, S. Zhang and W. Huang, Mater. Sci. Eng. A, 2005, 403, 25–31.
25. T. L. Barr, Appl. Surf. Sci., 1983, 15, 1–35.
26. T. Choudhury, S. Saied, J. Sullivan and A. Abbot, J. Phys. D: Appl. Phys., 1989, 22, 1185.
27. X.-F. Lu, D.-J. Wu, R.-Z. Li, Q. Li, S.-H. Ye, Y.-X. Tong and G.-R. Li, J. Mater. Chem. A, 2014, 2, 4706–4713.
28. Y.-C. Chiang, P.-C. Chiang and E.-E. Chang, J. Environ. Eng., 2001, 127, 54–62.
29. K. S. Sing and R. T. Williams, Adsorpt. Sci. Technol., 2004, 22, 773–782.
30. S. Zhu, J. Li, X. Deng, C. He, E. Liu, F. He, C. Shi and N. Zhao, Adv. Funct. Mater., 2017, 27, 1605017.
31. K. M. Khalil, J. Colloid Interface Sci., 2007, 307, 172–180.
32. W. Zhao, X. Li, R. Yin, L. Qian, X. Huang, H. Liu, J. Zhang, J. Wang, T. Ding and Z. Guo, Nanoscale, 2018, 11, 50–59.
33. L. Wang, M. Ara, K. Wadumesthrige, S. Salley and K. Y. S. Ng, J. Power Sources, 2013, 234, 8–15.
34. Y. Bae, Y. S. Yun, H.-D. Lim, H. Lee, Y.-J. Kim, J. Kim, H. Park, Y. Ko, S. Lee, H. J. Kwon, H. Kim, H.-T. Kim, D. Im and K. Kang, Chem. Mater., 2016, 28, 8160–8169.
35. D. Kundu, R. Black, E. J. Berg and L. F. Nazar, Energy Environ. Sci., 2015, 8, 1292–1298.
36. B. D. Adams, R. Black, C. Radtke, Z. Williams, B. L. Mehdi, N. D. Browning and L. F. Nazar, ACS Nano, 2014, 8, 12483–12493.
37. Y. Bae, D.-H. Ko, S. Lee, H.-D. Lim, Y.-J. Kim, H.-S. Shim, H. Park, Y. Ko, S. K. Park, H. J. Kwon, H. Kim, H.-T. Kim, Y.-S. Min, D. Im and K. Kang, Adv. Energy Mater., 2018, 8, 1702661.
38. E. Nasybulin, W. Xu, M. H. Engelhard, Z. Nie, S. D. Burton, L. Cosimbescu, M. E. Gross and J.-G. Zhang, J. Phys. Chem. C, 2013, 117, 2635–2645.
39. B. D. McCloskey, A. Valery, A. C. Luntz, S. R. Gowda, G. M. Wallraff, J. M. Garcia, T. Mori and L. E. Krupp, J. Phys. Chem. Lett., 2013, 4, 2989–2993.
40. X. Cao, X. Zheng, Z. Sun, C. Jin, J. Tian, S. Sun and R. Yang, Appl. Catal., B, 2019, 253, 317–322.
41. J. J. Xu, Z. W. Chang, Y. Wang, D. P. Liu, Y. Zhang and X. B. Zhang, Adv. Mater., 2016, 28, 9620–9628.
42. P. Zhang, S. Zhang, M. He, J. Lang, A. Ren, S. Xu and X. Yan, Adv. Sci., 2017, 4, 1700172.
43. L. Peiqing, X. Qunji and L. Weimin, Surf. Coat. Technol., 2001, 135, 118–125.
44. Y. Lu, H. Ang, Q. Yan and E. Fong, Chem. Mater., 2016, 28, 5743–5752.
45. J. Xu, L. Liu, P. Munroe, Z.-H. Xie and Z.-T. Jiang, J. Mater. Chem. A, 2013, 1, 10281.
46. P. F. Liu, S. Yang, L. R. Zheng, B. Zhang and H. G. Yang, Chem. Sci., 2017, 8, 3484–3488.

---

Footnote

† Electronic supplementary information (ESI) available: SEM, EDS, Raman, XPS, cycling performance, discharge/charge profile, EIS, rate performance, and TEM data. See DOI: 10.1039/c9ta10713d

---