Peng Zhang1,2, Shoufeng Zhang3, Mu He2,4, Junwei Lang1, Aimin Ren3, Shan Xu4, Xingbin Yan1. 1. Laboratory of Clean Energy Chemistry and Materials State Key Laboratory of Solid Lubrication Lanzhou Institute of Chemical Physics Chinese Academy of Sciences Lanzhou 730000 P. R. China. 2. University of Chinese Academy of Sciences Beijing 100039 P. R. China. 3. State Key Laboratory of Theoretical and Computational Chemistry Jilin University Jilin 130023 P. R. China. 4. State Key Laboratory for Oxo Synthesis and Selective Oxidation Lanzhou Institute of Chemical Physics Chinese Academy of Sciences Lanzhou 730000 P. R. China.
Abstract
Large Li2O2 aggregations can produce high-capacity of lithium oxygen (Li-O2) batteries, but the larger ones usually lead to less-efficient contact between Li2O2 and electrode materials. Herein, a hierarchical cathode architecture based on different discharge characteristics of α-MnO2 and Co3O4 is constructed, which can enable the embedded growth of large Li2O2 aggregations to solve this problem. Through experimental observations and first-principle calculations, it is found that α-MnO2 nanorod tends to form uniform Li2O2 particles due to its preferential Li+ adsorption and similar LiO2 adsorption energies of different crystal faces, whereas Co3O4 nanosheet tends to simultaneously generate Li2O2 film and Li2O2 nanosheets due to its preferential O2 adsorption and different LiO2 adsorption energies of varied crystal faces. Thus, the composite cathode architecture in which Co3O4 nanosheets are grown on α-MnO2 nanorods can exhibit extraordinary synergetic effects, i.e., α-MnO2 nanorods provide the initial nucleation sites for Li2O2 deposition while Co3O4 nanosheets provide dissolved LiO2 to promote the subsequent growth of Li2O2. Consequently, the composite cathode achieves the embedded growth of large Li2O2 aggregations and thus exhibits significantly improved specific capacity, rate capability, and cyclic stability compared with the single metal oxide electrode.
Large Li2O2 aggregations can produce high-capacity of lithium oxygen (Li-O2) batteries, but the larger ones usually lead to less-efficient contact between Li2O2 and electrode materials. Herein, a hierarchical cathode architecture based on different discharge characteristics of α-MnO2 and Co3O4 is constructed, which can enable the embedded growth of large Li2O2 aggregations to solve this problem. Through experimental observations and first-principle calculations, it is found that α-MnO2 nanorod tends to form uniform Li2O2 particles due to its preferential Li+ adsorption and similar LiO2 adsorption energies of different crystal faces, whereas Co3O4 nanosheet tends to simultaneously generate Li2O2 film and Li2O2 nanosheets due to its preferential O2 adsorption and different LiO2 adsorption energies of varied crystal faces. Thus, the composite cathode architecture in which Co3O4 nanosheets are grown on α-MnO2 nanorods can exhibit extraordinary synergetic effects, i.e., α-MnO2 nanorods provide the initial nucleation sites for Li2O2 deposition while Co3O4 nanosheets provide dissolved LiO2 to promote the subsequent growth of Li2O2. Consequently, the composite cathode achieves the embedded growth of large Li2O2 aggregations and thus exhibits significantly improved specific capacity, rate capability, and cyclic stability compared with the single metal oxide electrode.
Rechargeable lithium‐oxygen (Li‐O2) batteries are triggering worldwide interest due to their ultrahigh theoretical energy density (3505 Wh kg−1 based on lithium peroxide (Li2O2)), thus exhibiting significant potential to meet the demand of long‐range electric vehicles.1, 2 In a typical Li‐O2 battery, during the discharging process, O2 reacts with Li+ to form Li2O2 with insulating and insoluble characteristics that lead to the gradual increase of electrode impedance until the electron transport cannot match the current density.3, 4, 5 For this reason, if conformal Li2O2 film or small Li2O2 aggregation is formed, the electrode surface will be passivated early and thus yield a low capacity.6, 7 If large‐sized Li2O2 aggregation is produced, the passivation of electrode surface will be prolonged and result in a relatively high capacity.8, 9, 10 Therefore, the capacity of Li‐O2 batteries is strongly related to the size of Li2O2 aggregation. For instance, Bruce and co‐workers and Luntz and co‐workers found that the size of Li2O2 toroids could be adjusted by changing the solvent or electrolyte additive, and the battery capacity could be increased with Li2O2 sizes.6, 8 Han and co‐workers reported that Li2O2 with large sheet‐like morphology provided a higher capacity than the small Li2O2 nanoparticles.11 Amine and co‐workers also found that large Li2O2 toroids delivered higher capacity than Li2O2 film.12 Thus, a large size is a desirable feature for Li2O2 in Li‐O2 batteries.However, Chen and co‐workers and Byon and co‐workers both demonstrated that upsizing Li2O2 would result in a higher charging plateau and low charging rate.13, 14 That is, for large Li2O2 aggregations, a contradiction between high capacity and low oxygen evolution reaction (OER) overpotential is found. To solve this problem, many studies have focused on the development of OER catalysts. For example, Han co‐workers reported a PdCu/Super P cathode yielding large sheet‐like Li2O2 as well as low OER overpotential.11 Liu and co‐workers designed a carbon‐dotted/CoO/Super P cathode that could produce large Li2O2 toroids and perform lower overpotentials.15 Moreover, Peng and co‐workers have proved that the OER reaction interface of Li‐O2 batteries is electrode/Li2O2 instead of Li2O2/electrolyte.4 Basing on this result, further increasing contact sites between the catalyst and Li2O2, the decomposition of the large Li2O2 aggregations should become more easily and quickly. Meanwhile, large Li2O2 aggregations usually randomly deposit on the electrode surface,16, 17, 18, 19, 20 leading to the less effective contact between Li2O2 and electrode material. In this regard, confining the large Li2O2 aggregations in a hierarchical cathode/catalyst matrix should be an effective strategy to produce sufficient contact sites between them. Although many studies have focused on the construction of hierarchically structured cathodes,21, 22, 23, 24, 25 these reported structures still did not realize the strategy.For the construction of the cathode architecture toward the embedded growth of Li2O2, using metal oxides is one of the ideal choices due to their scientific and practical values compared with noble metals, and their variety and better stability compared with carbon materials.22, 26, 27, 28, 29, 30 Among various metal oxides, MnO2 and Co3O4 are the most studied materials,28, 31, 32, 33, 34, 35, 36 and the corresponding results have shown that MnO2 and Co3O4 are promising active materials for constructing ideal cathode architecture.Herein, we demonstrate that the embedded growth of large Li2O2 aggregations can be realized by constructing a hierarchical cathode architecture wherein Co3O4 nanosheets are grown on the surfaces of α‐MnO2 nanorods on a conductive carbon paper (CP). First, the different discharge characteristics of Co3O4 nanosheet and α‐MnO2 nanorod are analyzed through experimental observations and first‐principles calculations. α‐MnO2 nanorod tends to produce the uniform Li2O2 nucleation due to its preferential Li+ adsorption characteristic, resulting in the formation of uniform Li2O2 particles; whereas Co3O4 nanosheet tends to produce Li2O2 crystal seeds through the surface and solution due to its preferential oxygen adsorption characteristic, resulting in the separate formation of Li2O2 film and Li2O2 nanosheets. The composite cathode architecture wherein Co3O4 nanosheets are grown on α‐MnO2 nanorods exhibits extraordinary synergistic effects: α‐MnO2 can offer the initial Li2O2 nucleation sites and produce enough sites to grow ultrathin Co3O4 nanosheets, and Co3O4 nanosheets are conducive to yield dissolved LiO2. Thus, large mooncake‐like Li2O2 and sheet‐like Li2O2 both possessing embedded structures are formed on CP‐MnO2‐Co3O4 electrode at low and high current densities, respectively. As a consequence, such composite cathode exhibits remarkably improved electrochemical performance compared with the single ones, in terms of a large capacity (5950 mAh g−1 at 51 mA g−1), outstanding rate capability (2574 mAh g−1 at 1.03 A g−1), as well as good cyclic stability (54 cycles at the limited capacity of 1000 mAh g−1).
Results and Discussion
Structure Characterization of CP‐MnO2 and Its Discharge Products in Li‐O2 Cells
In this study, α‐MnO2 and Co3O4 were directly grown on CP substrates as freestanding electrodes to avoid using any binder that will increase the resistance and induce the possible side reactions in Li‐O2 cells.37, 38 Specifically, CP‐MnO2 was prepared via a facile hydrothermal method and its microstructure and morphology are shown in Figure
. X‐ray diffraction (XRD) pattern presented in Figure 1a reflects that the diffraction peaks can almost be well indexed to tetragonal α‐MnO2 (PDF#44‐0141) except the carbon peaks (26° and 53°). As shown in Figure 1b, scanning electron microscopy (SEM) images exhibit a uniform growth of highly ordered nanorods array on the CP substrate. The transmission electron microscopy (TEM) image further confirms the nanorod‐like morphology (Figure 1c) and the selected area electron diffraction (SEAD, Figure S1a, Supporting Information) suggests the single‐crystal structure of α‐MnO2 nanorod. Lattice fringes displayed in the high‐resolution TEM (HRTEM) are ≈0.5 and 0.27 nm, corresponding to the (200) and (101) planes, respectively. Combined with the Fourier transform image (inset of Figure 1c), the growth direction of α‐MnO2 nanorod can be confirmed to be at the [001] axis with the exposed planes of (020) and (110). When employed as the cathode for Li‐O2 battery, the CP‐MnO2 electrode delivers specific capacities of 2195, 1543, and 648 mAh g−1 at 52, 104, and 311 mA g−1, respectively. Correspondingly, the ORR (OER) voltage plateaus are ≈2.63 V (4.05 V), 2.52 V (4.15 V), and 2.40 V (4.11 V) at different currents (Figure 1d).
Figure 1
a) XRD pattern and b) SEM image of CP‐MnO2. c) TEM image of an individual MnO2 nanorod. Insets are the HRTEM image and its corresponding Fourier transform image. d) Charge–discharge curves of CP‐MnO2 at different currents. SEM images of CP‐MnO2 electrode discharged at e) 104 mA g−1 and f) 311 mA g−1. The insets in b), e), and f) are the corresponding magnified images.
a) XRD pattern and b) SEM image of CP‐MnO2. c) TEM image of an individual MnO2 nanorod. Insets are the HRTEM image and its corresponding Fourier transform image. d) Charge–discharge curves of CP‐MnO2 at different currents. SEM images of CP‐MnO2 electrode discharged at e) 104 mA g−1 and f) 311 mA g−1. The insets in b), e), and f) are the corresponding magnified images.As mentioned before, the morphology of Li2O2 can reflect the electrochemical performance of a Li‐O2 battery. Here, the discharge products were examined by SEM and XRD. Figure 1e,f shows CP‐MnO2 electrode morphology discharged at 104 and 311 mA g−1, respectively. The surfaces of MnO2 nanorods in two samples are both covered by numerous particles except for the different thicknesses. Here a thicker layer represents a higher capacity value. The XRD pattern of the discharged CP‐MnO2 exhibits the characteristic peaks of Li2O2 at about 32.7° and 35° and they disappear after recharging (Figure S2a, Supporting Information), indicating the reversibility of CP‐MnO2 electrode. The weak diffraction peaks of Li2O2 may be attributed to its low crystallinity and low content in this sample, which is also reflected by other metal oxide electrodes.39, 40, 41 Furthermore, the obvious peak shift of MnO2 to the low angle direction can be detected from the discharged electrode, suggesting the volume increase of the lattice.41, 42 The increased lattice volume should be attributed to the storage of LiO in 2 × 2 MnO6 octahedron tunnels of α‐MnO2 instead of Li+ insertion, which is due to the insertion of LiO clogging the channels and Li2O2 on the surfaces of MnO2 further covering the channels.31, 42, 43 After recharging, these peaks can shift back to their original positions, suggesting the reversible extraction of the LiO in the tunnels. Additionally, the electrochemical performance of the pure CP substrate was also investigated to confirm the major contribution of the loading active materials on battery capacity (Figure S3, Supporting Information).
Structure Characterization of CP‐Co3O4 and Its Discharge Products in Li‐O2 Cells
CP‐Co3O4 electrode was synthesized via a simple electro‐deposition method followed by an annealing process (Figure S4, Supporting Information). Its structure characteristics and electrochemical performance were analyzed and the corresponding results are presented in Figure
. As shown in Figure 2a, the major XRD peaks of CP‐Co3O4 sample are consistent with the pure Co3O4 phase (PDF#43‐1003) except for the diffraction peaks of CP. SEM images show the uniform and vertical growth of Co3O4 nanosheets on the CP skeleton, and these nanosheets interconnect into a 3D framework (Figure 2b). TEM images verify that such a Co3O4 nanosheet has an obvious surface‐porous structure (Figure 2c and Figure S5a, Supporting Information) and the nitrogen adsorption–desorption isotherms prove the large specific surface area of Cp‐Co3O4 sample (74.73 m2 g−1, Figure S6b, Supporting Information). The diffraction rings in the SEAD pattern (Figure S1b, Supporting Information) of the Co3O4 nanosheet indicate its polycrystalline structure, and the HRTEM and Fourier transform images (insets of Figure 2c) further show that the polycrystalline nanosheet is composed of numerous Co3O4 monocrystals. When employed as the cathode for a Li‐O2 battery, the CP‐Co3O4 electrode exhibits specific capacities of 2080, 1445, and 551 mAh g−1 at 51, 102, and 306 mA g−1, respectively (Figure 2d). These profiles also reveal that the discharge (charge) plateau voltages are 2.59 V (3.92 V), 2.60 V (3.98 V), and 2.27 V (3.95 V). The electrochemical properties shown here are similar to previously reported freestanding Co3O4 electrodes.44, 45, 46 XRD results of discharged and charged CP‐Co3O4 electrodes prove the reversible formation and decomposition of Li2O2 (Figure S2b, Supporting Information). Moreover, as displayed in Figure 2e, after the electrode discharged at 100 mA g−1, numerous Li2O2 nanosheets are generated on the electrode surfaces with a randomly floating state. Simultaneously, film‐like Li2O2 is also formed on the surfaces of Co3O4 nanosheets (seen from the inset). In comparison, when the current increases to 300 mA g−1, such nanosheet‐like morphology disappears from the electrode. Instead, film‐like Li2O2 and discrete nanoparticle aggregations occupy the electrode surface.
Figure 2
a) XRD pattern and b) SEM image of CP‐Co3O4. c) TEM image of an individual Co3O4 nanosheet. The insets are the HRTEM image and its corresponding Fourier transform image. d) Charge–discharge curves of CP‐Co3O4 electrode at different currents. SEM images of CP‐Co3O4 electrode discharged at e) 102 mA g−1 and f) 306 mA g−1. The insets in b), e), and f) are the corresponding magnified images.
a) XRD pattern and b) SEM image of CP‐Co3O4. c) TEM image of an individual Co3O4 nanosheet. The insets are the HRTEM image and its corresponding Fourier transform image. d) Charge–discharge curves of CP‐Co3O4 electrode at different currents. SEM images of CP‐Co3O4 electrode discharged at e) 102 mA g−1 and f) 306 mA g−1. The insets in b), e), and f) are the corresponding magnified images.
Discharge Characteristics of CP‐MnO2 and CP‐Co3O4
The morphology of the discharge product depends on the electrode material morphology, current density, and the intrinsic discharge characteristic of used electrode material. For the influence of electrode morphology, previous studies have reported that carbon nanotubes and granular KB carbon can both produce toroidal Li2O2. The morphology of electrode material is not the critical factor to determine its discharge characteristic.9, 47 Our recent study also reported that flower‐like Li2O2 assembled by Li2O2 sheets was both produced on Ni/Co3O4 nanowires and Ni/Co3O4 rectangular nanosheet electrodes.48 In this study, aside from setting the same mass current density for testing CP‐MnO2 and CP‐Co3O4 electrodes, we also compared the discharged morphology of the two electrodes at the same current density of ≈1.40 mA m−2
SSA based on the specific surface area (Table S1 and Figure S7, Supporting Information). As a result, the morphology of the discharged CP‐MnO2 electrode still exhibits uniform granular Li2O2 coating on α‐MnO2 nanorod surface. Thus, the intrinsic characteristic of metal oxides determines the growth process of Li2O2 in an operated Li‐O2 cell. Then, we deeply analyzed the different discharge characteristics of α‐MnO2 nanorod and Co3O4 nanosheet, respectively.Schematic illustrations of the discharging process of α‐MnO2 (Figure S8a, Supporting Information) and Co3O4 (Figure S8b, Supporting Information) and the corresponding equations were outlined. For α‐MnO2, the first ORR step was proven to be a Li+ adsorption process with desolvation (Equation (S1) in Figure S8 of the Supporting information),31, 32, 49 obtaining an electron and further adsorbing O2 to form LiO2* (“*” represents adsorbed species) at the MnO2 surface (Equation (S2) in Figure S8 of the Supporting information).50, 51 Then, LiO2* transforms to Li2O2 through an electrochemical reduction or a disproportionation reaction (Equation (S3) in Figure S8 of the Supporting information).50, 52 Owing to the single crystal structure of α‐MnO2 nanorods, the accommodation of Li2O2 in 2 × 2 MnO6 octahedron tunnels, and close LiO2 adsorption energy for exposed crystal face of α‐MnO2 nanorod (Figure
c,d),31, 42 the formed Li2O2 seeds can be uniformly distributed on the MnO2 surface and act as the desired seeds for subsequent Li2O2 growth. After the nucleation process of Li2O2 on α‐MnO2 surface, LiO2 can be adsorbed on the Li2O2 surface or dissolved into electrolyte due to the close adsorption energy of LiO2 on Li2O2 (−1.27 eV, Figure 3a) and the solvation energy of LiO2 (−1.35 eV, Figure 3b) (Equation (S4) in Figure S8 of the Supporting information). Thus, the uniform Li2O2 nucleus can induce the subsequent growth of Li2O2 simultaneously through surface and solution, finally resulting in the formation of Li2O2 particles (Equations (S5) and (S6) in Figure S8 of the Supporting information).
Figure 3
a) The adsorption energy of LiO2 on (1111) plane of Li2O2. b) The solvation energy of LiO2 in Tetraethylene glycol dimethyl ether (TEGDME). c,d) The adsorption energies of LiO2 on (110) and (020) planes of α‐MnO2. e–h) The adsorption energies of LiO2 on O‐(111), (112), (110), and (311) planes of Co3O4.
a) The adsorption energy of LiO2 on (1111) plane of Li2O2. b) The solvation energy of LiO2 in Tetraethylene glycol dimethyl ether (TEGDME). c,d) The adsorption energies of LiO2 on (110) and (020) planes of α‐MnO2. e–h) The adsorption energies of LiO2 on O‐(111), (112), (110), and (311) planes of Co3O4.For CP‐Co3O4 electrode, the first ORR step is oxygen adsorption followed by oxygen reduction to form O2
‐ on the electrode surface, and then O2
‐ transfers to electrolyte due to solvation (Equations (S7) and (S8) in Figure S8 of the Supporting information).10, 33, 53 The dissolved O2
‐
sol (“sol” represents solvated species) and Li+ form LiO2sol by coupling (Equation (S9) in Figure S8 of the Supporting information).54 Near the electrode surface, the coupled LiO2sol can be adsorbed on Co3O4 nanosheet surfaces owing to the larger adsorption energy between them (−3.10 to −4.63 eV on different crystal planes, Figure 3e–h) compared to the solvation energy of LiO2 in electrolyte (−1.35 eV, Equation (S10) in Figure S8 of the Supporting information). Then, the adsorbed LiO2* undergoes a second reduction or disproportionation to form Li2O2 as sur‐seeds (“sur‐seeds” represent the seeds formed through surface, Equation (S11) in Figure S8 of the Supporting information).55 Simultaneously, in the bulk electrolyte, LiO2sol transforms to Li2O2 through disproportionation (Equation (S12) in Figure S8 of the Supporting information).12 When Li2O2 cluster reaches a supersaturated state, solid Li2O2 deposits on the electrode surface as sol‐seeds (“sol‐seeds” represent the seeds formed through solution).6, 19 Thus, for Co3O4 material, the nucleation occurs simultaneously through surface and solution, which leads to further growth of Li2O2 on the surface and in the electrolyte.
Design Philosophy for the Composite Cathode
Basing on the above analyses, the nucleation processes of Li2O2 for α‐MnO2 and Co3O4 are quite different. α‐MnO2 undergoes Li+ adsorption and O2 reduction to directly form LiO2 on MnO2 surfaces. Co3O4 undergoes O2 adsorption, O2 reduction, O2
‐ solvation, coupling of O2
‐ and Li+, and then forms dissolved LiO2 oligomer or being adsorbed on electrode surfaces. Thus, the nucleation rate of α‐MnO2 should be faster than Co3O4 at the initial discharge process. Meanwhile, for Co3O4 electrode, the growth of Li2O2 through surface and electrolyte occurs at the same time. Taking these findings into account, the composite electrode, i.e., CP‐MnO2‐Co3O4 (Figure
e), may combine the discharge characteristics of α‐MnO2 and Co3O4 (Figure 4a–d), of which α‐MnO2 nanorods function as the initial growth sites of Li2O2 acting as the seeds for subsequent Li2O2 growth and simultaneously provide plenty of sites for Co3O4 nanosheets deposition inducing the formation of sufficient activity area to provide a large amount of dissolved LiO2. Then, as shown in Figure 4f, Li2O2 produced on the composite cathode architecture probably possesses a large size and an embedded structure, resulting in the remarkable improvement of battery performance.
Figure 4
e) Schematic illustration of designing the composite cathode architecture f) with ideal discharge product morphology based on different discharge characteristics of a,b) CP‐MnO2 and c,d) CP‐Co3O4.
e) Schematic illustration of designing the composite cathode architecture f) with ideal discharge product morphology based on different discharge characteristics of a,b) CP‐MnO2 and c,d) CP‐Co3O4.
Structure Characterization of CP‐MnO2‐Co3O4
As shown in Figure
a, the CP‐MnO2‐Co3O4 composite electrode was synthesized through electro‐depositing Co3O4 on the CP‐MnO2 substrate followed by an annealing process. Figure 5b displays the XRD pattern of as‐prepared composite electrode, presenting the diffraction peaks of both MnO2 and Co3O4. X‐ray photoelectron spectroscopy (XPS) survey also confirms the coexistence of MnO2 and Co3O4 (Figure 5c and Figure S9, Supporting Information). For Mn 2p spectrum, the Mn 2p3/2 (642.1 eV) and Mn 2p1/2 (653.8 eV) peaks have a spin‐energy separation of 11.7 eV, reflecting the +4 oxidation state for Mn.56 For Co 2p spectrum, peaks at 780.5 and 796.6 eV correspond to Co 2p3/2 and Co 2p1/2, respectively, indicating the characteristic of Co3O4 phase.57 Fitting using Gaussian method, the Co 2p spectrum can be well fitted into two spin orbits with features of Co2+ and Co3+. From the top view of SEM images, Co3O4 nanosheets interconnect each other to form numerous matrices for the accommodation of Li2O2 during the discharge process (Figure 5d and inset). From the side view (Figure S10, Supporting Information), Co3O4 nanosheets grow on the surface of each MnO2 nanorod from bottom to top, but these nanosheets do not cover the whole surface of MnO2 nanorods. The microstructure of Co3O4 nanosheets grown on a MnO2 nanorod was further revealed by TEM (Figure 5e and Figure S11, Supporting Information). As shown in a typical interface area marked with a blue circle in Figure 5e, the main lattice fringes spacing are 0.48 and 0.54 nm, corresponding to the (002) plane of MnO2 and the (110) plane of Co3O4, respectively. Moreover, as seen from the crystal planes of the Co3O4 nanosheet (their boundaries are marked with a dashed line), their crystal orientations are the same (marked with a red solid line). Nevertheless, the Co3O4 nanosheets in composited electrode still exhibit polycrystalline structure (Figure S12, Supporting Information).
Figure 5
a) Schematic of the preparation of CP‐MnO2‐Co3O4. b) XRD pattern of CP‐MnO2‐Co3O4. c) XPS Mn 2p and Co 2p spectra of CP‐MnO2‐Co3O4. d) SEM image of a single fiber in the composite electrode. The inset shows the layout of Co3O4 nanosheets on a MnO2 nanorod from the top view. e) TEM image of the interface between MnO2 nanorod and Co3O4 nanosheet.
a) Schematic of the preparation of CP‐MnO2‐Co3O4. b) XRD pattern of CP‐MnO2‐Co3O4. c) XPS Mn 2p and Co 2p spectra of CP‐MnO2‐Co3O4. d) SEM image of a single fiber in the composite electrode. The inset shows the layout of Co3O4 nanosheets on a MnO2 nanorod from the top view. e) TEM image of the interface between MnO2 nanorod and Co3O4 nanosheet.Furthermore, mesopores on Co3O4 nanosheet in the composite electrode are larger than those in CP‐Co3O4 electrode (Figure S5b, Supporting Information). The formation of these larger mesopores is due to much thinner Co3O4 nanosheets in CP‐MnO2‐Co3O4 sample compared with the ones in CP‐Co3O4 under the same annealing process (Figure S13, Supporting Information).58 With the consideration of the identical mass of Co3O4 on the two electrodes, the decreased thickness of Co3O4 nanosheets is linked to the significantly increased height derived from more depositing sites on the CP‐MnO2 substrate than those on the pure CP substrate (Figure S14, Supporting Information). As a result, the composite electrode exhibited a large surface area of 71.18 m2 g−1 (Figure S6c, Supporting Information), which would be in favor of producing a large amount of dissolved LiO2sol. Furthermore, the abundant mesoporous structure of the Co3O4 nanosheets may also facilitate the transport of electrolyte, oxygen, superoxide, and peroxide species. The surface area of the composite electrode is slightly lower than that of CP‐Co3O4 electrode (74.73 m2 g−1), mainly because of the low specific surface area (21.31 m2 g−1), high mass loading of MnO2 (0.85 mg cm−2 for MnO2; 0.51 mg cm−2 for Co3O4), larger mesoporous size of Co3O4, and partially overlapped surface of MnO2 and Co3O4 in the composite electrode.
Performance Comparison of CP‐MnO2, CP‐Co3O4 and CP‐MnO2‐Co3O4 Cathodes, and the Discharge Products of CP‐MnO2‐Co3O4 Electrode
The electrochemical performance of Li‐O2 battery with CP‐MnO2‐Co3O4 electrode was then investigated and the results are shown in Figure
. By comparing the charge–discharge curves of the three electrodes, it is clearly seen that CP‐MnO2‐Co3O4 electrode exhibits a much higher capacity (4850 mAh g−1) than the two other electrodes (CP‐MnO2 and CP‐Co3O4) at ≈103 mA g−1. SEM image of the discharged CP‐MnO2‐Co3O4 electrode was investigated. As expected, mooncake‐like Li2O2 is uniformly embedded in the given matrices (Figure 6b) and the formation of a large amount of Li2O2 aggregation endows the higher capacity of CP‐MnO2‐Co3O4 electrode. Such capacity improvement is more obvious at a high current density. As shown in Figure 6c, the composite electrode can deliver 3543 mAh g−1 at 309 mA g−1, almost six times higher than the two other electrodes. After discharging, numerous large Li2O2 sheets form on the surface of CP‐MnO2‐Co3O4 electrode with an inlaid structure (Figure 6d). Moreover, as shown in Figure 6e,f, the CP‐MnO2‐Co3O4 electrode exhibits superior rate capability. For instance, the discharge capacity is 3401 mAh g−1 at 618 mA g−1, 57.3% of the capacity at 51 mA g−1 (5940 mAh g−1). When the current increases to 1236 mA g−1, the capacity still reaches 2592 mAh g−1, 43.6% of the initial capacity. Additionally, we also compared the electrochemical performance of the three samples normalized by their electrode area, and the related results are shown in Table S2 and Figure S15 (Supporting Information). The results show that the performance improvement of the composite electrode is more remarkable when the test results are normalized by the electrode area, further reflecting the effectivity of this cathode architecture.
Figure 6
Charge–discharge curves of three electrodes at a) ≈103 mA g−1 and c) ≈309 mA g−1, respectively. SEM images of discharged CP‐MnO2‐Co3O4 electrode at b) 103 mA g−1 and d) 309 mA g−1, respectively. e) Rate capability of the Li‐O2 battery with CP‐MnO2‐Co3O4 electrode at different current densities. f) The capacity retention of CP‐MnO2‐Co3O4 electrode at different current densities.
Charge–discharge curves of three electrodes at a) ≈103 mA g−1 and c) ≈309 mA g−1, respectively. SEM images of discharged CP‐MnO2‐Co3O4 electrode at b) 103 mA g−1 and d) 309 mA g−1, respectively. e) Rate capability of the Li‐O2 battery with CP‐MnO2‐Co3O4 electrode at different current densities. f) The capacity retention of CP‐MnO2‐Co3O4 electrode at different current densities.We believe that the remarkably improved electrochemical performance should be attributed to the synergistic effect of the different discharge characteristics of MnO2 and Co3O4 through the scientific design of the cathode architecture. In detail, the obviously increased height of Co3O4 nanosheets in the vertical direction can significantly increase the amount of LiO2sol and boost the growth of Li2O2, resulting in the formation of large aggregations (mooncake‐like and large sheet‐like Li2O2). Simultaneously, these Li2O2 aggregations are embedded in the cathode architecture, which can increase the contact sites between electrode materials and Li2O2 with a result of reduced overpotential and significantly improved rate capability of Li‐O2 cell. Furthermore, during the discharge process, Li+ can intercalate into MnO2 crystal structure, as proven from the weaker peak shift in XRD pattern of discharged CP‐MnO2‐Co3O4 electrode than that in XRD pattern of discharged CP‐MnO2 electrode (Figures S16 and S2, Supporting Information). The decreased peak shift to the low angle indicates the reduced volume change during the discharge process, suggesting the reduced amount or shrinking volume of insert species. As seen from Figure S17 (Supporting Information), the close capacity values between CP‐MnO2 and CP‐MnO2‐Co3O4 tested in argon atmosphere indicate that Li+ can diffuse across the Co3O4 nanosheets (the specific capacity values of two electrodes were both calculated based on the mass of MnO2). Simultaneously, the O2 adsorption characteristic of Co3O4 can lead to the low availability of O2 on MnO2 surfaces. Thus, in this case, the inserted LiO is partially replaced by Li+ due to the overlapped interface that can allow Li+ diffusion but prevent O2 transmission. Fortunately, the lithiated MnO2 was proved to be a better electronic conductor than the pristine MnO2, facilitating the charge transfer during electrochemical reactions.41 Additionally, to confirm the key role of MnO2 in the composite structure, another composite electrode with a higher Co3O4 loading was prepared and its SEM images are presented in Figure S18a,b (Supporting Information). When the deposition quantity of Co3O4 increases from 1 to 2.5 C cm−2, the surface of MnO2 sample is completely covered by Co3O4 nanosheets. After discharging, this electrode can also form abundant Li2O2, resulting in a high discharge capacity (Figure S18c, Supporting Information). However, the discharge products do not embed in the matrices of the electrode, leading to a larger OER overpotential compared with the composite electrode with an appropriate Co3O4 loading (Figure S18c,d, Supporting Information).As shown in Figure
a, CP‐MnO2‐Co3O4 electrode exhibits the lowest overpotential among the three electrodes during the limited capacity test. Meanwhile, no obvious change exists in the cyclic charge–discharge profile (Figure 7b), reflecting its good cycle stability. In comparison, CP‐MnO2‐Co3O4 electrode is able to perform over 50 cycles before the terminal voltage decreases below 2 V, while CP‐MnO2 and CP‐Co3O4 electrodes can only achieve approximately ten cycles at a limited capacity of ≈1030 mAh g−1 (Figure 7c). Fourier Transform in frared (FTIR) spectra of the three electrodes at the eighth charge–discharge cycle were measured to investigate the reversibility of the three electrodes. As shown in Figure 7d, the obvious existence of the peaks around 400 to 600 cm−1 indicates the reversible formation and decomposition of Li2O2 for the CP‐MnO2‐Co3O4 composite electrode. For CP‐MnO2‐Co3O4 electrode, the weakest intensity of organic lithium salt peaks suggests the slightest electrolyte decomposition. Also, the peak of Li2CO3 at 870 cm−1 is observed for CP‐MnO2 and CP‐Co3O4 electrodes. The formation of Li2CO3 is mainly due to the oxidation of carbon during the charge process.59 Thus, this peak may derive from the side reaction related to CP substrate. As seen from FTIR spectra of CP‐MnO2‐Co3O4 electrode at the 30th cycle (Figure S19, Supporting Information), the reversible decomposition of Li2O2 can also be detected. However, the intensity of peaks for organic lithium salt increases at the same time, indicating the gradual accumulation of the byproducts on the composite cathode, which further leads to the passivated electrode surfaces and the limited cycle life (54 cycles for CP‐MnO2‐Co3O4 electrode).
Figure 7
a) Charge–discharge profiles of CP‐MnO2, CP‐Co3O4, and CP‐MnO2‐Co3O4 electrodes at ≈103 mA g−1 with a limited capacity of ≈1030 mAh g−1. b) Charge–discharge profiles of CP‐MnO2‐Co3O4 electrode at different cycles during cyclic tests. c) Cycling performance of the three electrodes measured at ≈103 mA g−1 under the limited capacity of ≈1030 mAh g−1. d) FTIR spectra of the three electrodes at the eighth cycle.
a) Charge–discharge profiles of CP‐MnO2, CP‐Co3O4, and CP‐MnO2‐Co3O4 electrodes at ≈103 mA g−1 with a limited capacity of ≈1030 mAh g−1. b) Charge–discharge profiles of CP‐MnO2‐Co3O4 electrode at different cycles during cyclic tests. c) Cycling performance of the three electrodes measured at ≈103 mA g−1 under the limited capacity of ≈1030 mAh g−1. d) FTIR spectra of the three electrodes at the eighth cycle.Because the discharge depth can also affect the cycling performance of Li‐O2 batteries, we further compared the cyclic stability of the three electrodes under the identical rate of limited discharge capacity to their full capacities. In detail, the capacity of CP‐MnO2‐Co3O4 electrode was 4850 mAh g−1 at 103 mA g−1 and 1030 mAh g−1 (≈21% of the full capacity) was chosen as the limited capacity during the cyclic tests. Here, this discharge ratio was also applied in CP‐MnO2 and CP‐Co3O4 electrodes (1543 mAh g−1 × 21% = 323 mAh g−1, with the value of 320 mAh g−1 chosen for the two electrodes). As shown in Figure S20 (Supporting Information), the cyclic stability of CP‐MnO2 and CP‐Co3O4 electrodes can be remarkably improved compared with that of the electrodes under the limited capacity setting at ≈1030 mAh g−1. Even so, their cycle lives are still inferior to CP‐MnO2‐Co3O4 electrode. Moreover, we compared the morphologies of the three electrodes after recharging (Figure S21, Supporting Information). α‐MnO2 nanorods are entirely covered by byproducts after being charging at the 25th cycle (may have some undecomposed Li2O2, Figure S21a, Supporting Information) and Co3O4 surfaces are partially covered by byproducts (Figure S21b, Supporting Information). Fortunately, no obvious byproduct was found on the charged CP‐MnO2‐Co3O4 electrode at 25th cycle (Figure S21c, Supporting Information). However, after recharging at the 54th cycle, undesired byproducts also covered on CP‐MnO2‐Co3O4 electrode surfaces (Figure S21d, Supporting Information). These results show that the composite electrode is conducive to alleviate the accumulation of byproducts, thus exhibiting enhanced cyclic stability. Consequently, when the composite electrode was tested at the limited capacity of 320 mAh g−1, it can achieve 142 cycles before the discharged terminal voltage decreases below 2 V (Figure S22, Supporting Information).
Conclusions
In summary, a cathode architecture composed of nanostructural α‐MnO2 and Co3O4 was developed to achieve embedded growth of discharge product Li2O2. To this end, the inherent catalytic characteristics of α‐MnO2 and Co3O4 are first studied through experimental observations and first‐principle calculations. For the single MnO2 and Co3O4, the CP‐MnO2 electrode produces Li2O2 with granular morphology due to its preferential Li+ adsorption property, accommodating LiO in 2 × 2 MnO6 octahedron channels and similar LiO2 adsorption energy of exposed crystal faces. The CP‐Co3O4 electrode produces nanosheet‐like Li2O2 as well as film‐like Li2O2 on the electrode surface owing to its preferential O2 adsorption characteristic and the difference in LiO2 adsorption energy of exposed crystal faces. On this basis, a hierarchical cathode architecture where Co3O4 nanosheets attach on MnO2 nanorods array was fabricated, and the as‐made composite electrode exhibits a distinct synergistic effect between α‐MnO2 and Co3O4. On the one hand, α‐MnO2 acts as the initial growing sites. On the other hand, ultrathin Co3O4 nanosheets with obviously increased height further produce a large amount of dissolved LiO2. Thus, uniform mooncake‐like Li2O2 and large sheet‐like Li2O2 can be generated on the composite electrode with an embedded structure at low and high current densities, respectively. As a consequence, this composite electrode can offer remarkably improved electrochemical performance when compared to the single α‐MnO2 or Co3O4 electrode, in terms of high reversible capacity, superior rate capability, and outstanding cycle stability. This study develops a feasible strategy to design reasonable cathode architectures for realizing the embedded growth of Li2O2. It also provides valuable information for understanding the discharge mechanism of Li‐O2 batteries and designing high‐performance Li‐O2 cathodes. Notably, this design philosophy may also be interesting for other energy storage devices.
Conflict of Interest
The authors declare no conflict of interest.SupplementaryClick here for additional data file.
Authors: Malachi Noked; Marshall A Schroeder; Alexander J Pearse; Gary W Rubloff; Sang Bok Lee Journal: J Phys Chem Lett Date: 2016-01-21 Impact factor: 6.475
Authors: B D McCloskey; A Speidel; R Scheffler; D C Miller; V Viswanathan; J S Hummelshøj; J K Nørskov; A C Luntz Journal: J Phys Chem Lett Date: 2012-03-30 Impact factor: 6.475
Authors: Jun Lu; Lei Cheng; Kah Chun Lau; Eric Tyo; Xiangyi Luo; Jianguo Wen; Dean Miller; Rajeev S Assary; Hsien-Hau Wang; Paul Redfern; Huiming Wu; Jin-Bum Park; Yang-Kook Sun; Stefan Vajda; Khalil Amine; Larry A Curtiss Journal: Nat Commun Date: 2014-09-12 Impact factor: 14.919
Authors: David G Kwabi; Vyacheslav S Bryantsev; Thomas P Batcho; Daniil M Itkis; Carl V Thompson; Yang Shao-Horn Journal: Angew Chem Int Ed Engl Date: 2016-01-28 Impact factor: 15.336