Zhengcai
Zhang‡
a,
Dulin
Huang‡
a,
Shuochao
Xing
a,
Minghui
Li
a,
Jing
Wu
a,
Zhang
Zhang
a,
Yaying
Dou
*ab and
Zhen
Zhou
*a
aInterdisciplinary Research Center for Sustainable Energy Science and Engineering (IRC4SE2), School of Chemical Engineering, Zhengzhou University, Zhengzhou 450001, China. E-mail: yydou@zzu.edu.cn; zhenzhou@zzu.edu.cn
bKey Laboratory of Advanced Energy Materials Chemistry (Ministry of Education), College of Chemistry, Nankai University, Tianjin 300071, China
First published on 22nd July 2024
Efficient catalysts are indispensable for overcoming the sluggish reaction kinetics and high overpotentials inherent in Li–O2 batteries. However, the lack of precise control over catalyst structures at the atomic level and limited understanding of the underlying catalytic mechanisms pose significant challenges to advancing catalyst technology. In this study, we propose the concept of precisely controlled pre-lithiated electrocatalysts, drawing inspiration from lithium electrochemistry. Our results demonstrate that Li+ intercalation induces lattice strain in RuO2 and modulates its electronic structure. These modifications promote electron transfer between catalysts and reaction intermediates, optimizing the adsorption behavior of Li–O intermediates. As a result, Li–O2 batteries employing Li0.52RuO2 exhibit ultrahigh energy efficiency, long lifespan, high discharge capacity, and excellent rate performance. This research offers valuable insights for the design and optimization of efficient electrocatalysts at the atomic level, paving the way for further advancements in Li–O2 battery technology.
Addressing these challenges, extensive research efforts have been directed towards both solid and liquid catalysts. However, the utilization of liquid catalysts often introduces the issue of the “shuttle effect”, resulting in the corrosion of lithium metal and subsequent reduction in battery durability.7 Alternatively, various solid catalysts, including metal oxides,8,9 alloys,5 and high-entropy catalysts,10 have been extensively investigated in Li–O2 batteries. Previous studies emphasized the importance of modulating the adsorption strength between LiO2 and catalyst surfaces to facilitate the formation and decomposition of Li2O2, a pivotal process in Li–O2 batteries.10–13
To further enhance the catalytic activity of these candidates, surface engineering techniques are commonly employed to modulate atom arrangement and electronic structure. Established methods such as crystal facet engineering, defect engineering, and surface/interface modification are widely utilized towards achieving these objectives.14–17 Among these strategies, doping with heteroatoms has shown promising electrocatalytic activity for boosting the performance of Li–O2 batteries. By introducing heteroatoms with varying valence states and electronegativity, the charge and spin density of materials can be redistributed, thereby influencing the adsorption of oxygen-containing intermediates at active sites.
For instance, research conducted by Lu et al. demonstrated that incorporating excess Co into the (101) plane of RuO2 results in abundant Ru/Co dual-atom sites on the RuO2 (110) surface. This approach effectively optimizes both charge transfer and the accessibility of the intermediate *OOH species in zinc–air batteries.18 However, traditional doping methods often entail complex preparation procedures, impeding precise control over the foreign element concentration and the rational design of catalysts. Moreover, the structure–activity relationship of catalysts prepared via traditional chemical methods, especially at the atomic level, remains elusive for oxygen electrochemical processes in Li–O2 batteries.
Therefore, it is imperative to develop an efficient and controllable preparation method that strikes a delicate balance between cost-effectiveness and the precise preparation of catalysts. This will provide a solid foundation for the development of high-performance Li–O2 batteries.
Electron-ion coupled transfer in electrochemistry offers a promising alternative for modifying the electronic or crystal structure of host materials. Unlike conventional chemical synthesis, electrochemical techniques operate at lower temperatures and pressures, leading to reduced energy consumption and waste generation. Furthermore, these methods afford greater control over impurity concentration through adjustable electrochemical parameters.19–21 This controllability allows for increased freedom in modulating the atom arrangement and electronic properties of catalytic materials, thereby facilitating the design and synthesis of tailored catalysts.
Electrochemical methods, including galvanic replacement, electrochemical exfoliation, and electrochemical insertion/extraction, have found wide application in the synthesis of energy catalytic materials, demonstrating promising outcomes in various electrocatalytic applications such as water splitting and carbon dioxide reduction.22–24 For example, the electrochemical treatment of Li2Co2O4 spinel facilitates the formation of amorphous active layers, thereby enhancing the oxygen evolution reaction (OER) due to the presence of Co4+ ions and oxygen sites with electronic holes.25 Similarly, treating MnO2 with lithium exhibited improved catalytic performance in Li–O2 batteries, indicating the promising recycling of depleted Li–MnO2 batteries.26 Although these examples underscore the distinct advantages of electrochemical treatments in fabricating efficient catalysts, the specific catalytic mechanisms still need to be further revealed, especially in Li–O2 batteries.
Drawing inspiration from this perspective, we propose a simple lithium electrochemical tuning method to enhance the catalytic activity of RuO2, the most commonly used representative in Li–O2 batteries. This method allows for the quantitative adjustment of Li+ concentration (x). The findings reveal that Li+ not only induces lattice strain by embedding into the lattice interstitial of RuO2 but also functions as an electron donor, directly modulating the electronic structure of RuO2. Specifically, the valence state of Ru decreases with Li+ intercalation, accompanied by the formation of oxygen vacancies. These modifications facilitate efficient electron transfer from the catalyst to the reaction intermediates while optimizing the adsorption behavior of the Li–O intermediates, particularly LiO2, on the electrode surface. Consequently, Li–O2 batteries employing Li0.52RuO2 as a catalyst demonstrate ultrahigh energy conversion efficiency and long-term reversibility. The elucidation of the atomic-level catalytic mechanism provides valuable insights into the rational design and optimization of advanced electrocatalysts for Li–O2 batteries.
To prepare LixRuO2, a Li+ intercalation process was employed. Specifically, the RuO2@CNT cathode was assembled into CR2032 coin cells, with Li foil as the counter-electrode. The electrolyte was a 1 M bis(trifluoromethane)sulfonamide lithium salt (LiTFSI) dissolved in tetraethylene glycol dimethyl ether (TEGDME). The Li+ intercalation into RuO2 was achieved by discharging the cell at a constant current density of 10 mA g−1 in an argon atmosphere. While the amount of Li+ intercalation was controlled by the discharge time. The obtained LixRuO2 was used as the as-prepared cathode for Li–O2 batteries.
Differential electrochemical mass spectrometry (DEMS) measurements of Li–O2 batteries. Quantitative DEMS was employed to investigate the stability and reversibility of Li–O2 batteries. A custom-designed Li–O2 battery, equipped with two securely attached poly(ether-ether-ketone) (PEEK) capillary tubes for gas inlet and outlet, was connected to a commercially available magnetic sector mass spectrometer (Thermo Fisher) using a specially engineered gas purging system. The flow rate of the purging gas was precisely regulated by a digital mass flow meter (Bronkhorst). During discharging, a gas mixture of Ar/O2 (mass ratio of 1:
4) with a controlled flux of 0.4 sccm served as the carrier gas to accurately measure the consumption of O2. For charging Li–O2 batteries, high-purity (99.999%) Ar was utilized as the carrier gas to quantify O2 evolution. The DEMS battery assembly and testing followed procedures similar to those outlined in the section on electrochemical measurements.
The differential charge density is calculated according to Δρ = ρAB − ρA − ρB, where ρAB, ρA, and ρB represent the charge densities of LixRuO2 (x = 0, 0.5) covered by Li2O2 with or without adsorbed LiO2, and isolated LiO2, respectively. Yellow and blue colors indicate the charge accumulation and depletion, respectively.
The adsorption energy is calculated according to the equation Eads = EAB − EA − EB, where EAB is the total energy of LixOy (x = 1, 2, 4, y = 2, 4) molecules adsorbed on the LixRuO2 (x = 0, 0.5), EA is the energy of isolated LixOy (x = 1, 2, 4, y = 2, 4) molecule, EB is the energy of LixRuO2 substrate.
XRD was conducted to investigate the influence of Li+ intercalation on the crystal structure of RuO2. Fig. S3† illustrates the XRD patterns of pristine RuO2, displaying three distinct diffraction peaks at approximately 28.1°, 35.1°, and 54.4°, in accordance with the characteristic diffraction pattern of rutile RuO2. The analysis of LixRuO2 primarily focused on the peaks at 35.1° and 54.4°, to circumvent the diffraction interference of carbon at 28.1°. Evidently, LixRuO2 retain the overall diffraction characteristics of RuO2. However, as the Li+ concentration increases, these peaks' positions gradually shift slightly towards lower angles, indicating expansion of the RuO2 lattice due to Li+ intercalation (Fig. 1b). The simulations, as depicted in Fig. S4,† further confirm this phenomenon. The lithium intercalation levels used were 0.08, 0.25, 0.33, and 0.5, based on experimental data and computational feasibility. The calculations illustrate that Li+ intercalates into the octahedral interstice formed by six adjacent O atoms, rather than replacing the Ru cations, thus leading to the expansion of the RuO2 lattice (Fig. 1c). The fitted lattice parameters of RuO2 before and after Li+ intercalation, along with the corresponding dilatation strains, are presented in Table S1.† Specifically, with increasing the lithiation degree, the expansion strains along a, b, and c-axis increase by 3.73%, 3.89%, and 0.26%, respectively.
The morphology and crystal structure of the-thus prepared samples were analyzed using SEM and TEM. As shown in Fig. S5,† the initial cathode exhibited a uniform distribution of RuO2 and CNT. After 16 h electrochemical treatment, no significant changes were observed, except for a slight reduction in pore size. Selected area electron diffraction (SAED) analyses shown in Fig. 2a reveal a series of lattice fringes in pristine RuO2, corresponding to the (110), (101), and (200) planes of the rutile RuO2. With Li+ intercalation, the lattice spacing gradually increases, as demonstrated in Fig. 2b–f. Specifically, in the case of Li0.52RuO2, the lattice spacings of the (110), (101), and (200) planes increase to 0.2427 nm, 0.2768 nm, and 0.3157 nm, respectively. This indicates that Li+ intercalation causes lattice expansion, aligning with the observed shift of characteristic peak positions towards lower angles in XRD.
![]() | ||
Fig. 2 The SAED patterns (a) of pristine RuO2 and HRTEM images (b–f) of the RuO2 with different Li+ concentrations. |
To investigate the influence of electrochemical treatment on surface chemical states and electronic structure of LixRuO2, XPS was employed. Fig. 3a illustrates the Ru 3d spectra of pristine RuO2 and a series of LixRuO2 samples. The Ru 3d5/2 spectrum exhibits peaks at 281 and 282 eV, corresponding to Ru(III) and Ru(IV), respectively. Two satellite peaks associated with Ru 3d3/2 are also observed. Pristine RuO2 exclusively displays the Ru(IV) peak at 282 eV, consistent with previous reports.35,36 However, upon Li+ intercalation, a characteristic peak of Ru(III) at 281 eV emerges. The content of Ru(IV) and Ru(III) in LixRuO2 samples is summarized in Fig. 3c. With increasing Li+ concentration, the proportion of Ru(IV) decreases to 93.5%, 87.3%, 76.2%, and 71.4%. Moreover, both the Ru(IV) and Ru(III) peaks exhibit a slight shift towards higher binding energies, indicating electron transfer and a decrease in electron density around the Ru sites. Furthermore, an analysis of O 1s spectra was conducted, as presented in Fig. 3b. The peak around 529.4 eV in RuO2, attributed to lattice oxygen, gradually shifts to lower binding energies with Li+ intercalation. This shift suggests an increased electron density surrounding the oxygen sites, indicating partial electron transfer from the Ru sites to the O sites facilitated by Li+ intercalation.37,38 Notably, after 8 h intercalation, a new peak appears around 530.5 eV, corresponding to oxygen vacancy, whose proportion gradually increases with prolonged intercalation time, as statistically demonstrated in Fig. 3d. This increase might provide additional active sites for oxygen electrochemical reactions.
![]() | ||
Fig. 3 XPS of Ru 3d (a) and O 1s (b) of RuO2 and LixRuO2, and (c and d) content comparison of different chemical species calculated from the fitted XPS spectra. |
Fig. 4c shows the first-cycle charge–discharge curve of Li–O2 batteries with RuO2 or LixRuO2 cathode, which is another crucial criterion for evaluating the catalytic activity of materials. Compared with the RuO2 cathode, Li–O2 batteries based on LixRuO2 demonstrated smaller charge overpotentials, which is negatively correlated with the Li+ concentration. That is, the higher the Li+ intercalation level, the lower the reaction overpotential. Notably, the first charge voltage of Li0.52RuO2-based Li–O2 batteries decreased to approximately 3.41 V, which could effectively mitigate parasitic reactions at higher voltages. Moreover, Li–O2 batteries based on RuO2 displayed a limited cycle life of 150 cycles, while the batteries incorporating LixRuO2 demonstrated significantly improved cycling performance (Fig. 4d). Among them, the Li–O2 batteries utilizing Li0.52RuO2 exhibited the best cycling stability, with a remarkable cycle life of 300 cycles and stable operation exceeding 3000 hours (Fig. 4e). These results indicate that the Li+ intercalation significantly enhances the catalytic activity of LixRuO2 towards Li2O2 decomposition, which can be attributed to additional vacancy oxygen and the modulated electronic structure, providing additional active sites and enhancing reaction kinetics.
The practical feasibility of Li0.52RuO2-based Li–O2 batteries was evaluated at an increased current density of 500 mA g−1. Fig. 5a and b illustrate that the Li–O2 batteries with Li0.52RuO2 demonstrated improved cycling life of 321 cycles, surpassing that with untreated RuO2. Even at a higher cutoff capacity of 1000 mA h g−1, as displayed in Fig. 5c, Li–O2 batteries with Li0.52RuO2 demonstrated remarkable cycling stability for 153 cycles, whereas those with RuO2 showed limited cycle life of 111 cycles due to rapid voltage increase (Fig. 5d). These results highlight the exceptional ability of LixRuO2 to mitigate charging voltage and enhance cycling stability, demonstrating its practical potential. The rate performance of Li–O2 batteries utilizing Li0.52RuO2 is depicted in Fig. 5k. In comparison with pristine RuO2, the Li0.52RuO2-incorporating battery exhibits minimal discharge and charge voltage fluctuation, even under a high current density of 1000 mA g−1, which can be attributed to enhanced kinetics of oxygen electrochemical reactions. These findings underscore the paramount significance of modulating the atom structure and electronic feature of the catalyst in enhancing the performance of Li–O2 batteries.
To gain deeper insights into the underlying catalytic mechanism of LixRuO2, which is closely linked to the component and morphology of Li–O2 battery products and their electrochemical performance, XRD and SEM were employed to examine the cathodes in different discharge/charge states. As demonstrated in Fig. S6,† both the discharged RuO2 and LixRuO2 cathodes exhibited diffraction peaks at 32.9° and 35.0°, corresponding to the (100) and (101) crystal planes of Li2O2 (PDF#09-0355), indicating Li2O2 as the primary discharge product. Upon charging completion, the Li2O2 diffraction peak vanished, while the cathode peak reappeared, suggesting complete decomposition of discharge products for both RuO2 and LixRuO2 cathodes. Considering that XRD analysis provides information solely on the crystalline components, it is crucial to employ quantitative techniques like DEMS to evaluate the reversibility of Li–O2 batteries.39 The amount of O2 consumption/evolution during battery operation can be monitored using DEMS, which is imperative for the assessment of truly rechargeable Li–O2 batteries.
For an ideally reversible Li–O2 battery, the ratio of electrons to O2 molecule (e−/O2) shall be 2.0, and O2 is the only gaseous species involved in the discharge/recharge cycle. The typical galvanostatic discharge/charge profiles and the corresponding gas consumption/evolution rate are shown in Fig. 6. For the RuO2 based Li–O2 battery (Fig. 6a), a significantly deviated value of 2.39 e−/O2 was obtained upon discharge, with an ORR efficiency of only 80.5%, suggesting much undesired parasitic reaction. However, the e−/O2 ratio was quantified to be 2.08 (≈2.0 e−/O2) for the Li0.52RuO2-based Li–O2 battery, as depicted in Fig. 6b, with a slight deviation of 4% from the theoretical value. This negligible discrepancy could be attributed to the inevitable shuttle effect of oxygen and Li–O intermediates. The results indicate that the discharge reaction, catalyzed by Li0.52RuO2, primarily involved Li2O2 formation, which is consistent with the XRD result. Furthermore, the catalytic activity of Li0.52RuO2 and RuO2 during recharge was also evaluated using DEMS. As exhibited in Fig. 6c, the RuO2 based Li–O2 battery displays a high charge potential and a widely observed OER profile with a dip in the middle of charge, which usually is accompanied by a hydrogen evolution reaction (HER) resulted from the 1O2 attack, mirroring the missing O2 in the OER profile.40 Additionally, a significant amount of CO2 was observed when the charging voltage reached approximately 4.0 V. Actually, the appearance of gaseous CO2 during recharge is an indicator that the Li–O2 batteries are not ideally reversible, and the amount of CO2 generated directly reflects the extent of undesired parasitic reactions, which has been suggested to originate from the decomposition of carbon cathodes or electrolytes. On the contrary, the battery charged with Li0.52RuO2 (Fig. 6d) does not display a dip in its OER profile, which demonstrated a continuous, stable release of O2, with negligible CO2 generation. As a result, the ratio of charge passed to O2 evolved with Li0.52RuO2 cathode was quantified to be 2.11, which is much lower than 2.26 of the RuO2-based Li–O2 battery. Based on these findings, the disappearance of these three features (OER dip, CO2 release and ratio of e−/O2), the Li0.52RuO2 further confirms its superior catalytic activity. Besides, the parasitic products also were investigated through XPS. After the first cycle, the RuO2 cathode exhibited undecomposed Li2O2 and significant amounts of Li2CO3 byproducts (Fig. S7a†), likely due to elevated charging voltage. These byproducts, due to their wide band gap, are difficult to decompose during cycling, leading to increased charging voltage and eventual cathode passivation. As shown in Fig. S7b,† more Li2CO3 accumulated in the RuO2 cathode surface after the 10th cycle. In contrast, pre-lithiated cathodes exhibited significantly lower Li2CO3 levels after cycling. The Li2CO3 content decreased progressively with increasing Li+ concentration. Notably, the Li0.52RuO2 cathode showed almost no Li2CO3 byproducts, consistent with DEMS results. Even after the 10th cycle, no significant Li2CO3 was observed, indicating the system's exceptional capability in suppressing side reactions. Therefore, the incorporation of Li0.52RuO2 in Li–O2 batteries not only improves reaction kinetics but also reduces charging voltage, leading to reduced side reactions and enhanced reversibility, thereby improving the overall cycle stability.
![]() | ||
Fig. 6 DEMS analyses of gas consumption (a and b) and evolution (c and d) during discharge/charge of Li–O2 batteries based on (a and c) RuO2 and (b and d) Li0.52RuO2 cathodes. |
SEM was conducted to investigate the morphological features of discharge products. As depicted in Fig. S4a,† the pristine RuO2 electrode exhibited a homogeneous mixture of RuO2 particles and CNTs. Upon discharge, the electrode was covered by a dense film-like discharge product (Fig. S8a†), which may be the potential reason for limited discharge capacity. However, in addition to the dense film-like discharge product, LixRuO2 also exhibited some rod-like products, as presented in Fig. S8b–e.† Furthermore, increasing Li+ concentration promoted the growth of these products, corresponding to higher discharge capacity. This result is closely linked to the adsorption behavior of reaction intermediates on the catalyst surface, which is modulated by the electronic structure resulting from Li+ intercalation. It may optimize the adsorption strength of LixRuO2 cathodes toward the superoxide intermediates, promoting different oxygen reduction reaction (ORR) routes. Notably, the Li0.52RuO2 cathode exhibited the highest discharge product yield and capacity. Upon charging, residual discharge products were observed on RuO2 cathodes (Fig. S8f†), which severely reduce the availability of active sites and impede electron transfer. Conversely, the LixRuO2 cathode displayed complete products decomposition, showcasing excellent reversibility (Fig. S8g–k†). These results emphasize the importance of electrochemically modulating the electronic structure of materials, optimizing the adsorption characteristics of catalysts towards intermediate species in Li–O2 batteries. Further detailed explanations will be provided in DFT calculations sections.
Furthermore, the PDOS in RuO2 and LixRuO2 were recorded to reveal the regulating effect of Li+ insertion on the d-band center of RuO2. As shown in Fig. 7a, the d-band center of LixRuO2 exhibits a significantly upshift from −1.86 (x = 0) to 1.62 eV (x = 0.5), approaching the Fermi level. Meanwhile, the adsorption energy of LixRuO2 toward the key LiO2 intermediate gradually increase −3.02 eV to −3.85 eV, as depicted in Fig. 7b. Moreover, the adsorption profiles of other Li–O intermediates on RuO2 and Li0.5RuO2 were calculated, as illustrated in Fig. S10.† It reveals that the adsorption energy of all Li–O intermediates on the RuO2 (110) plane is significantly lower than that on the Li0.5RuO2 (110) plane, indicating that the incorporation of Li+ strengthens the interaction between Li–O intermediates and the catalyst. Notably, the strong binding interaction, particularly between LiO2 and Li0.5RuO2 cathodes, assumes a pivotal role in determining the growth route of discharge products and facilitating the OER catalytic activities.6,45,46 Visualizations of the differential charge density distributions (Fig. S11†) provide further support for the enhanced adsorption of catalysts towards the reaction species following Li+ intercalation. The electron donation and accumulation between O and the catalyst surface was presented by color of cyan and yellow, respectively. Remarkably, there are fewer electrons transferred from Ru to O on the RuO2 surface compared to the Li0.5RuO2 (110) surface. Based on the aforementioned calculation results, Fig. 7c presents a schematic illustrating the enhanced catalytic performance of RuO2 with Li+ insertion. Specifically, the remarkable improvement in electron transfer ability and adsorption functionality could synergistically optimize the reaction pathways and kinetics of the ORR and OER in Li–O2 batteries.
Integrating computational calculation with experimental results, we have elucidated the crucial role of the pre-lithiation RuO2 catalyst in promoting the nucleation and growth of Li2O2. Typically, upon ORR, dissolved oxygen initially undergoes a one-electron reduction process, forming LiO2 intermediate. For the RuO2 cathode with weak adsorption, a large number of soluble intermediates were formed at the initial stage of discharge and captured by porous electrodes. As the discharge process advanced, the intermediates distributed uniform crystal seeds into the porous structure, and finally induced the growth of film-shaped Li2O2 in accordance with SEM observation.47 However, for the LixRuO2 cathode, Li2O2 growth occurs through dual growth pathways with distinct morphologies. Specifically, a film-like Li2O2 similar to that on the RuO2 cathode is formed on the CNT surface. Additionally, due to the high affinity between LiO2 and LixRuO2 configurations, a significant confinement effect leads to the formation of rod-like Li2O2 products.11 Furthermore, the charge density distribution shown in Fig. S12† indicates that even when the Li0.5RuO2 surface is covered by Li2O2, Li0.5RuO2 still exhibits strong interactions with LiO2. Consequently, Li0.5RuO2-based batteries can sustain discharge, resulting in a larger discharge capacity. During the subsequent charging process, the enhanced interaction between Li0.5RuO2 and LiO2 intermediates, as well as Li2O2 products, facilitate the charge transfer between oxygen-containing species and oxygen electrode, thereby the OER kinetics. As a result, the discharge products can be decomposed at an ultra-low charging potential, while preventing the accumulation of residue from the discharge process and ensuring the ultralong cycle life for Li–O2 batteries.
Footnotes |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4sc03242j |
‡ Zhengcai Zhang and Dulin Huang contributed equally to this work. |
This journal is © The Royal Society of Chemistry 2024 |