Ultrathin Mesoporous Co3O4 Nanosheet Arrays for High-Performance Lithium-Ion Batteries.

Transition metal oxides, such as Co3O4, have attracted great attention for lithium-ion batteries (LIBs) due to their high theoretical capacity and satisfactory chemical stability. However, the slow kinetics of Li-ion and electron transport as well as poor cycling stability still largely restrains their applications. Here, we report the rational design of well-defined mesoporous ultrathin Co3O4 nanosheet arrays (NSAs) by topological transformation of layered double hydroxides nanosheet arrays (NSAs), which demonstrate significantly enhanced performance as anode for LIBs. The as-obtained Co3O4 NSAs with suitable thickness and abundant mesopores show excellent electrochemistry performance for LIBs, giving a high specific charge capacity of 2019.6 mAh g-1 at 0.1 A g-1, a good rate capability, and a remarkable cycling stability (1576.9 mAh g-1 after the 80th cycle), which is much superior to that of Co3O4 with thicker or thinner nanosheets as well as to that of the reported results. This facile strategy may be extended to the synthesis of other transition metal oxide NSAs, which can be potentially used in energy storage and conversion devices.

Introduction

Nowadays, rechargeable batteries have gradually become indispensable for advanced electrochemical energy storage systems, such as portable electronic devices and electric vehicles.[1−3] Typically, lithium-ion batteries (LIBs) have long been employed as an attractive power source due to their low cost, high energy density, and satisfactory cycling stability.[4,5] In practice, the energy output of LIBs is usually determined by their electrode materials, and in particular, one of the main challenges in the investigation of LIBs is to increase the capacity of anode materials.[6−9] To date, extensive attempts have been focused on the advanced anode by rational materials design, including construction of mesoporous architectures or conductive channels. For instance, hierarchical materials with mesoporous nanostructures can significantly improve the electrochemical performance of LIBs owing to their abundant exposure to active sites and increased transport of Li ions.[10−12] Although much progress has been made, developing anode materials with large energy density and long endurance is still essential for high-performance LIBs.

Transition metal oxides have been investigated as promising anode materials for LIBs.[13−15] Among various types of transition metal oxides, Co3O4 has attracted considerable attention due to their cost effectiveness and high theoretical capacity. However, the practical application of Co3O4 is still largely restrained by the slow kinetics of Li-ion and electron transport as well as poor cycling stability (large volume change and severe particle aggregation).[16,17] Recently, two-dimensional (2D) ultrathin materials, with the advantages of high surface-to-mass ratio and unique physicochemical properties, have drawn increasing research interest in electrochemical energy field. Considering the merits of 2D materials, if ultrathin Co3O4 nanosheets can be constructed by virtue of material exploration and fabrication strategy, a promising electrochemical performance would be achieved due to highly exposed surface atoms for faradic redox reactions. In addition, an elaborate ultrathin architecture also shortens the diffusion pathway of lithium ions, which guarantess a fast mass/electron transport.

Herein, the well-defined mesoporous ultrathin Co3O4 nanosheet arrays (NSAs) have been successfully fabricated via topological transformation of layered double hydroxide (LDH) precursor, which shows significantly enhanced performance as anode for LIBs. The CoIICoIII–LDH NSAs with various thicknesses were first obtained via a facile electrochemical synthesis method,[18] which thereby can be transformed to well-distributed mesoporous Co3O4 NSAs with ultrathin thickness ranging from 2 to 10 nm. Moreover, a “thickness-dependent LIBs performance” has been demonstrated by using Co3O4 NSAs as anode. Co3O4 NSAs with a thickness of 3 nm give the best performance with a maximum specific capacity (2019.6 mAh g–1) and satisfactory stability. It is worth mentioning that the optimal specific capacity of ultrathin Co3O4 NSAs is relatively high compared to that of Co3O4 anodes ever reported. In addition, this facile strategy can be extended to the synthesis of other ultrathin transition metal oxide NAs with potential application in energy storage and conversion.

Results and Discussion

Well-defined mesoporous ultrathin Co3O4 NSAs were fabricated through a topological transformation process (Figure A). Various CoIICoIII–LDH NSAs with different thicknesses on nickel foam substrates were first synthesized by using a rapid electrosynthesis method, which was further transformed to ultrathin Co3O4 NSAs via subsequent calcination treatment at 450 °C in air for 2 h. Figure B illustrates the X-ray diffraction (XRD) patterns of the CoIICoIII–LDH-T2 and Co3O4-T2 samples on the Ni foil substrate. CoIICoIII–LDH NSAs show a typical LDH phase with clear reflection peaks at 11.7, 23.5, 34.8, 59.7, and 60.9°, corresponding to the (003), (006), (012), (110), and (113) reflections of CoIICoIII–LDH. After calcination, the original LDH peaks are replaced by a series of new reflections at 19.0, 31.3, 36.8, 59.4, and 65.2°, corresponding to the (111), (220), (311), (511), and (440) reflections of Co3O4.[19−21] This indicates the successful transition from CoIICoIII–LDH to Co3O4. In addition, XRD patterns of CoIICoIII–LDH-T1, -T3 and Co3O4-T1, -T3 are shown in Figure S1, which show the same peak position with the corresponding precursors CoIICoIII–LDH-T2 and Co3O4-T2, respectively. In addition, the very weak reflection peak at 43.3° and relatively smooth surface (Figures S2 and S3) demonstrated that just limited NiO was formed from the Ni substrate under this heating condition. Figure C illustrates the Fourier transform infrared (FT-IR) spectra of CoIICoIII–LDH and Co3O4 powder collected from Ni foil substrate. The broad intense band at about 3450 cm–1 for CoIICoIII–LDH is due to the OH stretching mode of the hydroxyl groups in the host layers and the interlayer water molecules. The strong bands at 1383 and 662 cm–1 correspond to the absorption peak of intercalated nitrate anions and Co–O vibration of CoIICoIII–LDH. After calcination, the characteristic peaks of CoIICoIII–LDH disappear; two strong absorption bands at 660 and 567 cm–1 are observed, corresponding to the absorption peak of Co–O vibration of Co3O4.[22]

Figure 1

(A) Schematic illustration for the synthesis of Co3O4 NSAs with different thicknesses; (B) XRD patterns and (C) FT-IR spectra of CoIICoIII–LDH-T2 and Co3O4 NSAs-T2.

The morphologies of CoIICoIII–LDH and the obtained Co3O4 NSAs were investigated by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). It was observed that the thickness of the as-obtained CoIICoIII–LDH decreased from 10 to 2 nm along with the decrease of cobalt nitrate concentration in the electrosynthesized electrolyte (denoted CoIICoIII–LDH-T1, -T2, and -T3, respectively; Figure A–C). Figure D–F shows the SEM images of the corresponding Co3O4 NSAs (denoted Co3O4-T1, -T2, and -T3). It can be observed that the original vertically aligned nanosheet nanostructure is well-maintained for the Co3O4-T1 (Figure D) and Co3O4-T2 (Figure E). However, the nanoplatelet structure is destroyed when the LDH precursor is too thin (Co3O4-T3; Figure F). The TEM images (Figure G–I) for the obtained Co3O4 NSAs demonstrate the presence of uniform mesopores for the sample with medium thickness (Co3O4-T2), probably resulting from the dehydration and phase transformation of LDH nanosheets. In contrast, the Co3O4 NSAs with thicker nanosheets (Co3O4-T1) display limited mesopores, whereas the thinner Co3O4 (Co3O4-T3) NSAs give larger pores due to the serious collapse of the sheetlike nanostructure. The Co3O4 NSAs with different thicknesses and porous structures were further investigated by atomic force microscopy (AFM). As shown in Figure A, the sample of Co3O4-T1 NSAs shows a thickness of ∼10 nm, consistent with the TEM results. After decreasing the thickness of LDH precursor, the thickness of the obtained Co3O4 nanosheet obviously decreases to 3 nm for Co3O4-T2 NSAs (Figure B) and to 2 nm for Co3O4-T3 NSAs (Figure C). Compared to that of Co3O4-T1 NSA and Co3O4-T3 NSA samples, the Co3O4-T2 NSA sample displays a more ordered ultrathin platelike structure with uniform pores.

Figure 2

(A–C) SEM images of CoIICoIII–LDH with different thicknesses; SEM and TEM images of (D, G) Co3O4-T1, (E, H) Co3O4-T2, and (F, I) Co3O4-T3.

Figure 3

AFM images with cross-sectional results for (A) Co3O4-T1, (B) Co3O4-T2, and (C) Co3O4-T3; (D) high-resolution transmission electron microscopy (HRTEM) image and selected area electron diffraction (SAED); and (E) scanning transmission electron microscopy (STEM) image and (F) corresponding energy-dispersive X-ray spectrometry (EDS) mapping results of Co3O4 NSAs.

The detailed crystal structure and composition of the ultrathin CoIICoIII–LDH and Co3O4 NSAs were characterized by HRTEM and scanning transmission electron microscopy (STEM). Figure S4A shows the ultrathin LDH nanosheet structure, and the d-spacing from high resolution (Figure S4B) is about 0.25 nm, corresponding to the LDH (012) plane, which is consistent with the XRD pattern of CoIICoIII–LDH-T2.

As shown in STEM image (Figure D), the (111) and (311) planes with well-resolved lattice fringes of 0.467 and 0.24 nm are identified, indicating the high purity of the obtained Co3O4 NSAs. The same results can also be observed from the selected area electron diffraction (SAED) image, which shows three clear quasi-single-crystalline patterns with d-spacings of 0.467, 0.28, and 0.24 nm, corresponding to the (111), (220), and (311) planes of Co3O4 (Figure D, inset). Scanning transmission electron microscopy (STEM) image of Co3O4 NSAs and the corresponding energy-dispersive X-ray spectrometry (EDS) mapping verify the well-distributed elements of cobalt and oxygen (Figure E,F).

X-ray photoelectron spectroscopy (XPS) was used to demonstrate the composition and valence state of the precursor CoIICoIII–LDH-T2 and as-obtained Co3O4-T2 sample. As shown in Figure A, the full XPS spectra of CoIICoIII–LDH-T2 and Co3O4-T2 NSAs reveal signals of O 1s and Co 2p at 531.4 and 782.6 eV, respectively, indicating the existence of oxygen and cobalt elements in both samples. The XPS data of CoIICoIII–LDH-T1, -T3 and Co3O4-T1, -T3 are shown in Figures S5 and S6, which show nearly the same peak position with CoIICoIII–LDH-T2 and Co3O4-T2, respectively. The high-resolution Co 2p XPS spectra (Figure B, bottom curves) of CoIICoIII–LDH-T2 show two types of Co spaces at 796.5 and 797.6 eV, which can be attributed to Co2+ and Co3+. A similar result can also be observed on the Co 2p3/2 XPS spectrum (the peaks at 782.5 and 780.8 eV can be attributed to Co2+ and Co3+, respectively), indicating the presence of both Co2+ and Co3+ oxidation states in CoIICoIII–LDH NSAs-T2. After calcination of CoIICoIII–LDH NSAs-T2, both Co 2p3/2 and Co 2p1/2 peaks shift to a lower energy level: Co 2p1/2 peak moves to 796 eV (Co2+) and 794.6 eV (Co3+) (Figure B, top curves), and Co 2p3/2 peak moves to 782.5 eV (Co2+) and 780.8 eV (Co3+). The O 1s spectrum (Figure S6) displays two peaks at 531.4 and 532.1 eV for CoIICoIII–LDH, but shifts to 529 and 530 eV for Co3O4, which indicate the formation of Co3O4.[23,24]

Figure 4

(A) Full survey of XPS spectra for CoIICoIII–LDH-T2 and Co3O4 NSAs-T2; (B) XPS of Co 2p peaks of CoIICoIII–LDH-T2 and Co3O4 NSAs-T2.

Figure 5

(A) Cyclic voltammetries (CVs) of Co3O4 measured at a voltage range of 0.005–3.0 V versus Li, at a scan rate of 0.1 mV s–1. (B) Potential–capacity curves of Co3O4-T2 at 0.1 A g–1. (C) Charge–discharge profiles of Co3O4-T2 at different current densities and (D) rate performance of Co3O4 NSAs at different current densities from 0.1 to 2 A g–1 in the voltage range of 0.005–3.0 V. (E) Cycling performance of Co3O4 NSAs at 0.1 A g–1 rates for 80 cycles.

The electrochemical performance of the well-defined mesoporous ultrathin Co3O4 NSAs as an anode material for LIB was further studied. The CV curve for the Co3O4-T2 NSAs was first carried out at a scan rate of 0.1 mV s–1 with the potential window of 0.005–3.0 V. As shown in Figure A, two cathodic peaks appear in the first cycle at ∼0.5 and 0.8 V for Co3O4-T2 NSAs and disappear in the subsequent cycles, which are attributed to the formation of solid electrolyte interphase (SEI) film.[25] During the anodic scan, a peak located at ∼2.2 V is observed, which can be ascribed to the oxidation of metallic Co to Co3O4 and the decomposition of Li2O. The reversible progress of charge and discharge can be expressed as follows[26]In the second cycle, the reduction peak shifts to a higher potential at ∼1.0 V, whereas the oxidation peak position is almost unchanged. In the further subsequent cycles, the CV curves show good reproducibility and similar shapes, suggesting a high reversibility of lithium storage.[27]Figure B demonstrates the potential–capacity curves tested at a current rate of 0.1 A g–1. The irreversible capacity in the initial discharge process can be ascribed to the formation of SEI film, which is consistent with the CV results (Figure A). In contrast, the charge–discharge curves remain almost coincident in the subsequent cycles, indicating a good reversibility and cycle stability for the Co3O4 NSAs. It is worth noting that Co3O4-T1 and Co3O4-T3 NSAs (Figure S7A,C) exhibit similar voltage plateaus compared to that of Co3O4-T2 in the charge and discharge curves but give a clear lower specific capacity. This demonstrates that the thickness of Co3O4 NSAs plays a key role in determining their electrochemical properties. In contrast, the CV curve for the pristine Ni foam substrate after calcination is also explored (Figure S8A). It is clear that the Ni foam has almost no capacity contribution, and the calcined substrate just gives limited capacity (<5% of the value of Co3O4 NSAs supported on Ni foam), which is consistent with the result of potential–capacity curves (Figure S8B).

The rate capability and cycling performance of the as-obtained Co3O4 NSAs under different current densities were determined. The charge–discharge platforms for all of the samples at different current densities are similar (Figures C and S7B,D), indicating the same electrochemical pathways. Figure D shows the rate capability of Co3O4 NSAs with different thicknesses. The Co3O4-T2 NSAs give discharge capacities of 2019.6, 1904.6, 1622.3, 1176.7, and 632.7 mAh g–1 at 0.1, 0.2, 0.5, 1, and 2 A g–1, respectively. Along with the current density going back to 0.1 A g–1 after several cycles, the capacity of the Co3O4-T2 NSAs recovers to 1753.8 mAh g–1, exhibiting promising rate capability. In contrast, the Co3O4-T1 NSAs and Co3O4-T3 NSAs display relatively lower capabilities of 1261.1 and 1762 mAh g–1, respectively, at 0.1 A g–1. These values just remain at 987.4 and 1523.7 mAh g–1, respectively, when back to 0.1 A g–1 after a few cycles, which further indicates that the suitable thickness of Co3O4 anode can effectively facilitate Li+ insertion/extraction processes. The bulk Co3O4 with a thickness of ∼40 nm has also been synthesized (Figure S9), which gives discharge capacities of 814.6, 759.2, 537, 333, and 159 mAh g–1 at 0.1, 0.2, 0.5, 1, and 2 A g–1, respectively. Along with the current density going back to 0.1 A g–1 after several cycles, the capacity of bulk Co3O4 recovers to 566.45 mAh g–1, which shows obvious poor rate performance compared to that of mesoporous ultrathin Co3O4 nanosheets. The Co3O4 NSAs also show superior initial and reversible capacities compared to those of the reported Co3O4 anodes, to the best of our knowledge (Figure S10).[28−36] It is worth mentioning that the theoretical capacity of Co3O4 has been reported to be 890 mAh g–1, which is obviously less than that of the synthesized Co3O4 NSAs. This similar result has also been reported previously, where the capacities of Co3O4 ranged from 970 to 1615.8 mAh g–1. The clearly improved capacity compared with the theoretical values can be ascribed to the varied nanostructures, which change the Li+ storage sites and capacities.[31]

The cycling performance of the Co3O4 NSAs at a current density of 0.1 A g–1 is shown in Figure E. The first irreversible capacity loss can be ascribed to the formation of the SEI film during the first discharge. From the second cycle, the Coulombic efficiencies for all of the Co3O4 NSA samples reach nearly 100%. Moreover, the Co3O4-T2 NSAs retain a remarkable reversible capacity of 1576.9 mAh g–1 at the end of the 80th cycle, corresponding to a capacity decay of 0.28% per cycle. However, the Co3O4-T1 and Co3O4-T3 NSAs give 0.65 and 0.70% loss of capacity per cycle, respectively. The cycling stability of the Co3O4 NSAs at a higher current rate (1.0 A g–1) is further studied (Figure S11). The capacities of Co3O4-T1, -T2 and -T3 have reached a steady state after 200 cycles at 1 A g−1, which maintains 31.7%, 38.5%, 37% of their initial capacities, respectively, after 500 cycles. This demonstrates that the Co3O4 with ultrathin thickness also displays a higher capacity and improved stability at a high current. The independent nanosheet structure would relieve agglomeration and collapse during the cycle. Moreover, the porous structure provides more active sites and an expanded volume during the reaction.[37] For further understanding, we disassembled the battery after the 80th cycle and observed the morphology of the Co3O4 NSAs by SEM (Figure S12). We found that the nanosheet array structure of the Co3O4-T2 NSAs is still maintained, which proves that suitable thickness of the nanosheet with abundant mesopores suppresses structure collapse and aggregation during the charging/discharging process.

To further understand the improved lithium storage performance of Co3O4 NSAs, the electrochemical surface area (ECSA) of Co3O4 with different thicknesses was studied. The ECSA provides information of active surface areas, which can be determined by using double-layer capacitance (Cdl) (Figure S13). It is hard to determine the ECSA of Co3O4 in the battery electrolyte. Alternatively, we obtained this value in the 1 M KOH solution, which also can reveal the trend of the active site exposure for the Co3O4 with different nanostructures. It was found that the Cdl of Co3O4 NSAs increases first along with the increase of thickness from 2 to 10 nm, but it decreases by further increasing the NSA thickness (Figure A). This indicates that a suitable thickness of Co3O4 NSAs improves their electrochemical surface area and results in enhanced Li+ insertion/extraction. Moreover, the electrochemical impedance spectroscopy (EIS) spectra of Co3O4 provide additional information about the charge transport properties. As shown in Figure B, the Nyquist plots for all of the Co3O4 NSAs possess a semicircle in the high-frequency region and a straight line in the low-frequency region. The obviously depressed semicircle for the Co3O4-T2 NSAs indicates a decreased charge transport resistance. The increased slope of the Co3O4 NSAs with medium thickness also implies the enhanced Li+ diffusion in the electrode. As reported in previous works,[38−41] the total capacity of an electrode can be divided into three charge-storage mechanisms: Li+ insertion process, pseudocapacitance effect, and double-layer effect. The degree of capacitive effect can be indicated by analyzing the relationship between measured current (i) and scan rate (ν) from the CV data. The equations are listed as follows:When the b-value is 0.5, the charge-storage mechanism is considered as the diffusion process; however, when the b-value is 1.0, capacitive response is the main charge-storage mechanism. Figure S14B show CVs of Co3O4-T2 at different scanning rates; the calculated b-values are 0.56 and 0.535 (close to 0.5) at 1.0 and 1.1 V during the cathodic process (Figure C), indicating a main charge-storage mechanism of the diffusion process. In other potential window, the b-values range from 0.7 to 1.0, which implies that a part of capacity comes from pseudocapacitive effects.

Figure 6

(A) Linear plot of capacitive current and scan rate. (B) EIS spectra for Co3O4-T1, -T2, -T3 NSAs. (C) Calculated b-values as a function of voltage for cathodic and anodic sweeps for Co3O4-T2 NSAs. (D) Relationship between peak current and the square root of scan rate. (E) Schematic illustration for the discharge process of Co3O4 NSAs with different thicknesses (left: thick, inefficient utilization; right: thin, efficient utilization).

To further clarify the effects of thickness on the ionic transport properties, the apparent diffusion coefficient of Li-ion (DLi) was evaluated for Co3O4-T1, -T2, -T3 NSAs (see the Supporting Information for detailed calculations). The peak current demonstrates a linear relationship with the square root of scan rate (Figures D and S14). The different slopes of the Co3O4 NSAs imply the varied ability of Li+ diffusion in the electrode. According to the relationship of the peak current and the CV sweep rate, DLi of Co3O4-T1, -T2, -T3 NSAs are calculated to be 2.99 × 10–10, 5.44 × 10–10, and 1.67 × 10–10 cm2 s–1. The Co3O4-T2 NSAs show the largest ion diffusion coefficient, demonstrating a thickness-dependent Li+ storage property. Thus, the Co3O4 NSAs exhibit significantly enhanced electrochemical properties, which can be ascribed to the shortened lithium-ion transport distance within the nanosheets as well as fully utilized active sites during charging/discharging processes (Figure E).

Conclusions

In summary, ultrathin mesoporous Co3O4 NSAs were successfully prepared by calcining the precursor CoIICoIII–LDH NSAs, which showed an ordered sheetlike nanostructure and uniform mesopore distribution. The as-obtained Co3O4-T2 NSAs with suitable ultrathin thickness and abundant mesopores show an excellent electrochemistry performance for LIBs, giving a high specific charge capacity of 2019.6 mAh g–1 at 0.1 A g–1, a good rate capability, and a remarkable cycling stability (1576.9 mAh g–1 after the 80th cycle). This superior property for Li+ storage is attributed to the ultrathin nanosheet structure, which gives a large specific surface area and shortens the transport pathway of lithium ions. This work provides a facile and effective route for the preparation of ultrathin metal oxide NSAs for LIBs, which can be extended to other applications in energy storage and conversion systems.

Experimental Section

Reagents and Materials

All chemicals purchased were used without further purification. Co(NO3)2·6H2O and KOH were purchased from Aldrich Ltd. (Shanghai, China). Dehydrated ethanol, acetone, and HCl were purchased from Beijing Chemical Corp. The deionized water used in all experiments was purified through a Millipore system.

Synthesis of CoIICoIII–LDH NSA Precursor

The CoIICoIII–LDH NSAs were synthesized by a fast electrosynthesis method. First, a nickel foam (diameter: 1.5 cm) substrate was pretreated with acetone, 3 M HCl solution, deionized water, and ethanol, which was used as the working electrode and placed in an electrochemical cell in a three-electrode configuration (Pt wire as the counter electrode and saturated calomel electrode (SCE) as the reference electrode). The dissolved Co(NO3)2·6H2O in 50 mL of deionized water was used as electrolyte. The CoIICoIII–LDH NSAs grown on the surface of Ni foam were then prepared by a potentiostatic deposition process at a potential of −1.0 V versus SCE. Different thicknesses of CoIICoIII–LDH were obtained by changing the concentration of Co(NO3)2·6H2O to 0.15, 0.075, and 0.05 M (named CoIICoIII–LDH-T1, T2, and T3, respectively).

Preparation of Ultrathin Co3O4 NSAs

The ultrathin Co3O4 NSAs were obtained by calcining the as-synthesized CoIICoIII–LDH NSAs in air atmosphere. The typical heating process was carried out in a muffle furnace, heating from room temperature to 450 °C at 5 °C min–1, and then maintained for 2 h. Finally, the product (Co3O4 NSAs) was slowly cooled down to room temperature. The mass loading for all of the ultrathin Co3O4 NSAs is about 1.0 mg cm–2 by controlling the electrosynthesis time.

Structural Characterizations

X-ray diffraction patterns of the CoIICoIII–LDH and Co3O4 NSAs were collected on a Shimadzu XRD-6000 diffractometer using a Cu Kα source, with a scan range between 3 and 70° at a scanning speed of 10 °C min–1. FT-IR spectra were recorded on a NICOLET NEXUS470 Fourier transform infrared spectrometer. X-ray photoelectron spectra (XPS) were performed on a Thermo VG ESCALAB 250 X-ray photoelectron spectrometer at a pressure of about 2 × 10–9 Pa using Al Kα X-rays as the excitation source. The binding energies obtained in the XPS analysis were calibrated for specimen charging by referencing the C 1s peak to 284.80 eV. The morphologies of CoIICoIII–LDH and Co3O4 were investigated using a scanning electron microscope (SEM; Zeiss SUPRA 55) with an accelerating voltage of 20 kV, combined with energy-dispersive X-ray spectroscopy (EDS) for the determination of metal composition. Transmission electron microscopy (TEM) images were recorded with Philips Tecnai 20 and JEOL JEM-2010 high-resolution transmission electron microscopes. The accelerating voltage was 200 kV in each case. The thickness of Co3O4 was investigated by atomic force microscopy (AFM) (Multimode Nanoscope IIIa, Veeco Instruments) in tapping mode.

Electrochemical Test

The obtained materials were first cut into disc shapes (15 mm in diameter) by a slicer. Then, the electrodes were tableted into thin sections (pressure control of 5–10 MPa). The cointype (2025) cells were assembled in an argon-filled glovebox using pure lithium as the counter electrode, 1 M LiPF6 (ethylene carbonate/dimethylcarbonate 1:1 in volume) as the electrolyte, and a microporous membrane (Celguard 2400) as the separator. Cyclic voltammetry (CV), electrochemical surface area (ECSA), and electrochemical impedance spectroscopy (EIS) were carried out on a CHI660E electrochemical workstation (Shanghai Chenhua Instrument Co, China). The cells were galvanostatically charged and discharged between 0.005 and 3.0 V versus Li+/Li on a LAND test system at room temperature. The double-layer capacitance (Cdl) of Co3O4-T1, -T2, -T3 was calculated on the basis of the CV curves in 1.0 M KOH, which was measured in a non-faradaic region of 0.15–0.25 V at various scan rates. Charging current density differences (ΔJ = Ja – Jc) measured at 0.2 V were plotted against scan rate (Ja and Jc are the anodic and cathodic current densities, respectively, and the linear slope is twice that of Cdl).