Enhanced Interfacial Kinetics and High Rate Performance of LiCoO2 Thin-Film Electrodes by Al Doping and In Situ Al2O3 Coating.

The structure and surface-interface instability of LiCoO2 thin-film electrodes during charge-discharge cycles are one of the main factors leading to the deterioration of electrochemical performance. Element doping and surface coating are effective strategies to tackle this issue. In this work, Al-doped and in situ Al2O3-coated LiCoO2 composite thin-film electrodes are prepared by magnetron sputtering. The results show that the resultant composite thin-film electrodes exhibited excellent cycling stability, with a discharge specific capacity of 40.2 μAh um-1 cm-2 after 240 cycles at 2.5 μA cm-2, with a capacity retention rate of 94.14%, compared to a discharge capacity of the unmodified sample of only 37.7 μAh um-1 cm-2 after 110 cycles, with a capacity retention rate of 80.04%. In addition, the rate performance of the prepared LiCoO2 film is significantly improved, and the discharge specific capacity of the Al-doped sample reaches 43.5 μAh um-1 cm-2 at 100 μA cm-2, which is 38.97% higher than that of the unmodified sample (31.3 μAh um-1 cm-2). The enhancement of electrochemical performance is mainly attributed to the synergistic effect of Al doping and in situ Al2O3 coating. The metal Al forms a conductive network in the film, while part of the Al will enter the LiCoO2 lattice to form a LiAl y Co1-y O2 solid solution, promoting the transport of lithium ions and improving the stability of the electrode structure. The in situ continuous deposition of the coating optimizes the active material coating-electrolyte interface.

Introduction

With the development of technology, microelectronic devices such as flexible electronics, wearable smart devices, and portable electronics are being used increasingly widely.[1−7] Microelectronic devices are limited in size, and their energy storage devices need to be adapted to them in terms of space and structure.[8,9] The all-solid-state thin-film lithium battery is widely regarded for its high energy density, small size, safety, and structural designability, which can better meet the above requirements.[10−14] However, its application is limited by its thin electrode and low active material leading to low areal energy density and small overall capacity. Increasing the thickness of the active electrode (mainly the cathode) is an important way to improve the areal energy density and capacity of the thin-film battery. However, with increasing electrode thickness, the transport path length of lithium ions and electrons within the film increases, resulting in a decrease in transport capacity and ultimately a decrease in battery capacity and rate performance.[15] Taking the LiCoO2 film as an example, when the film thickness reaches 1 μm, the rate performance of the battery will decrease, and when the thickness reaches 4 μm, the capacity can only be released by 70%.[16] To avoid the above problems, researchers have developed a three-dimensional structure (3D) battery, which can increase the battery capacity and increase the areal energy density without increasing the thickness of the electrode film.[17,18] However, 3D batteries involve complex processes, high technical requirements, and poor mechanical properties. In addition, the contact interface between the electrode and electrolyte also has an important impact on the battery. Wang et al.[16] found that there is a certain thickness of interfacial disorder layer between the cathode film and the solid electrolyte, and its thickness will keep increasing with the increase of operating temperature and the progress of cycling, which will make the interfacial impedance increase continuously due to the lack of electrochemical activity of the interfacial disorder layer, resulting in the deterioration of the battery performance.

For the above problems of thin-film electrodes and their interfaces with electrolytes, the main solutions include the following: (1) Developing new electrode materials or designing new battery structures. For example, using high-throughput methods to prepare new positive and negative electrode materials based on the material genome is a systematic and complex project.[17] (2) Preparation of a composite thin film electrode and use of conductive materials (Ag,[18] C,[19] etc.) in the composite to improve the conductivity of the electrode. For example, Lee et al.[20] used magnetron sputtering to prepare LiCoO2/Ag/LiCoO2/Ag/LiCoO2 multilayer structure films. Because the presence of Ag improves the electronic conductivity of the films, the rate characteristics of the composite films are significantly better than those of LiCoO2 single-layer films. However, the Ag layer hinders the transport of lithium ions to a certain extent; that is, the conductivity of the thin film electrode can be improved by compounding with conductive materials, but it is difficult to improve the transport capacity of electrons and lithium ions in the electrode at the same time, so the modification effect is limited. (3) Element doping.[21] The lattice structure of the active material can be stabilized by element doping, and the cycle stability of the battery can be improved. Xie et al.[22] used ALD technology to dope a LiCoO2 cathode with Al. Since the existence of Al stabilizes the electrode structure and promotes the transport of lithium ions, the high-potential (4.7 V) cycle performance and rate performance of the electrode are significantly improved after doping. However, this work is only focused on LiCoO2 powder particles, and there is a lack of systematic research on LiCoO2 thin films. (4) The crystallographic orientation of the film is controlled to optimize the transport path of lithium ions and electrons.[23,24] Dai[25] prepared LiCoO2 films with different selective orientations using magnetron sputtering, investigated the electrochemical properties of films with different crystalline orientations and found that it is difficult to jointly optimize the capacity and cycling performance of single-phase LiCoO2 films. (5) Electrode surface interface treatment.[26−29] The electrode–electrolyte interface is stabilized by depositing a coating layer on the electrode surface to reduce the interface impedance.[30] It was reported that Al2O3 coating can significantly improve the electrochemical performance of LCO films.[31,32] Woo et al.[33] coated Al2O3 on the surface of a LiCoO2 positive electrode and found that the Al2O3 coating layer can effectively inhibit the interdiffusion between LiCoO2 and the solid electrolyte and reduce the increase in interfacial impedance during cycling. Further research found that the Al2O3 coating layer can provide a channel for the transport of lithium ions on the surface of the active electrode and improve the diffusion coefficient of lithium ions.[34,35] However, the ex situ surface coating of thin-film electrodes will form a relatively obvious physical and chemical interface between the electrode and coating layer. During the continuous intercalation and removal of lithium ions, due to the inconsistent volume change between the electrode and coating layer, cracks and peeling can easily occur in the coating interface, resulting in an increase in battery resistance and deterioration in performance.

In this paper, Al-doped LiCoO2 composite film was prepared by multitarget cosputtering technology to construct the transport channels of lithium ions and electrons to improve their transport capacity. Furthermore, the electrode surface was optimized by reactive sputtering with an Al2O3 coating deposited in situ on the surface of the Al-doped LiCoO2 composite film, thereby obtaining a high-performance thin-film cathode. The effect of Al doping on the electrical conductivity and structure of the thin film electrodes was analyzed, as well as the effect of the Al2O3 coating on the interfacial properties of LiCoO2 and the electrolyte.

Experimental Section

Preparation of Thin-Film Electrodes

The Li1.2CoO2 target (99.8% purity, purchased from Zhongnuo New Materials Co., Ltd.) and Al target (99.999% purity, purchased from Zhongnuo New Materials Co., Ltd.) were sputtered using magnetron sputtering equipment to prepare the related thin-film electrodes (schematic in Figure ). The Li1.2CoO2 and Al targets were mounted in the sputtering chamber of the magnetron sputtering equipment, and the background vacuum was controlled below 1.0 × 10–4 Pa. The clean substrate was mounted in the substrate holder and transferred to the sputtering chamber through the sample chamber. The substrate was heated and rotated (10 rpm), heated to the experimental temperature, and kept for 30 min. The working gas was introduced to start the sputtering. The sputtering time was controlled so that the thickness of the prepared film was approximately 500 nm, and the sample was cooled freely in the sputtering chamber after sputtering. (1) Preparation of LiCoO2 film: sputtering of the Li1.2CoO2 target alone with the following parameters: substrate heating temperature of 800 °C, sputtering power of 200 W, target base distance of 9.5 cm, working air pressure of 2 Pa (pure Ar), with the obtained LiCoO2 film recorded as LCO. (2) Preparation of Al-doped LiCoO2 films: Al-doped LiCoO2 films were prepared by cosputtering an Al target and a Li1.2CoO2 target. The spacing of the Al target from the substrate was 11.5 cm, and the sputtering power of the Al target was 5, 10, and 20 W (corresponding samples were marked as LCO-A5, LCO-A10, and LCO-A20). The Li1.2CoO2 target and other preparation parameters were the same as those used for the preparation of LCO films. (3) Preparation of the in situ composite film: After the sputtering deposition of the LCO-A10 film, the Li1.2CoO2 target was closed, the Ar gas inlet valve was closed, O2 gas was passed, the working air pressure was adjusted to 0.5 Pa, the Al target sputtering power was adjusted to 80 W, sputtering was continued to prepare an in situ Al2O3-coated Al-doped LiCoO2 film (labeled as LCO-A10@Al2O3), and the thickness of the Al2O3 coating layer was approximately 10 nm.

Figure 1

Schematic of the magnetron sputtering.

Characterization of Physicochemical Properties of Materials

A high-temperature X-ray diffractometer (XRD, X’pert PRO MPD, Panaco, The Netherlands) was used for the physical phase analysis of the material. Scanning electron microscopy (SEM, ZEISS Gemini 300) was used to observe the surface microscopic morphology of the samples as well as the particle shape. The materials were analyzed for composition using energy spectrometry (EDS, ZEISS Gemini 300), X-ray energy spectrometry (XPS, Thermo Kalpha Thermo ESCALAB 250XI), and Raman spectrometry (Thermo SCIENTIFIC ESCALAB 250Xi). The electrical conductivity of the films was tested using a four-probe resistivity tester (Suzhou Jingle Electronics Co., Ltd., RTS-8).

Battery Assembly and Electrochemical Performance Testing

The LiCoO2 film electrode prepared above was used as the cathode, polypropylene microporous (Celgard 2400) was used as the separator, Li foil (Guangdong Canrd New Energy Technology Co., Ltd.) was used as the anode, and 1 M LiPF6 in ethylene carbonate/ethyl methyl carbonate/dimethyl carbonate (EC/EMC/DMC, 1:1:1 by vol %) was used as the liquid electrolyte and assembled into a CR2032 coin cell. A LAND CT2001A system was used to test the charge–discharge and rate performance of the battery at 3.0–4.2 V (vs Li/Li+). Electrochemical impedance spectroscopy (EIS) measurements (frequency range 1.0 × 10–2 to 1.0 × 105 Hz, with an AC amplitude of 5 mV) and cyclic voltammetry (CV) with different sweep speeds (0.05 to 0.6 mV s–1) were carried out by using electrochemical workstations (Shanghai Chenhua, CHI1040C, CHI650E).

Results and Discussion

Al-Doped LiCoO2 Thin Films

Figure shows the SEM images of the Al-doped LiCoO2 film. It can be observed from the figure that the LCO contains obvious crystal particles and a loose and porous surface, which is beneficial to increasing the contact area between the electrode and electrolyte but also leads to the occurrence of more serious surface side reactions.[36] Compared with the undoped sample, the surface of the Al-doped LiCoO2 thin film becomes flat and dense, showing a certain degree of crystallinity. With an increase in the Al doping amount, the surface density of the film increases, which is beneficial to reducing the anisotropy and maintaining the electrode structure and surface interface stability. In addition, with an increase in the doping amount, the crystalline particles of the film gradually decrease, which may be due to the existence of Al which affects the nucleation and growth of LiCoO2 particles.[37]

Figure 2

SEM images of (a) LCO, (b) LCO-A5, (c) LCO-A10, and (d) LCO-A20

In order to further investigate the effect of cosputtered aluminum on the structure of LiCoO2 thin films, we performed XRD tests on the prepared samples, and the results are shown in Figure S1. The results show that the prepared LiCoO2 films show strong (003) direction and weak (101) peaks, and the intensity of the (003) peak gradually decreases with the increase of Al injection power, indicating that the crystallinity of the films gradually becomes worse, and the crystallinity is the worst when the Al sputtering power is 20 W, which is consistent with the SEM test results. Compared with the LCO samples, the (003) peaks of the doped samples are all shifted to a small angle. The reason is that the LiCo1–AlO2 solid solution formed by partial Al doping into the LiCoO2 lattice during the cosputtering process expands the c-axis.[37] During the charging and discharging process, the increased layer spacing along the c-axis will provide better migration channels for the lithium ions.[38]

To investigate the doping amount of Al during cosputtering, we performed EDS tests for the undoped and doped LiCoO2 films, and the results are shown in Figure . Figure a shows the energy spectrum and element content table for the LCO thin film sample. It can be observed from the figure that the atomic percentages (at %) of Co and O are 33.18 and 66.82 at %, respectively, and the ratios are very close to 1:2, which is the same as the atomic ratio of the elements in the chemical formula of LiCoO2. It can be observed from Figure b–d that the Co/O atomic ratio of the Al-doped sample is also very close to 1:2, which satisfies the element atomic ratio of the chemical formula of LiCoO2. An obvious Al element peak is observed in the samples after Al doping, indicating that Al successfully enters the interior of the LiCoO2 film, and with increasing Al target sputtering power, the amount of Al doping gradually increases. The atomic number ratios for Al doping in the LCO-A5, LCO-A10, and LCO-A20 samples are 0.15, 0.62, and 0.96 at. %, respectively. In addition, we found that the Co/O ratio in the samples increased with increasing Al content, which may be due to the loss of lithium and oxygen due to excessive doping, a result that is consistent with the XRD test results.[21]

Figure 3

EDS images of (a) LCO, (b) LCO-A5, (c) LCO-A10, and (d) LCO-A20.

Galvanostatic charge–discharge tests were carried out to investigate the effect of Al doping on the electrochemical properties of LiCoO2 thin-film electrodes. The discharge curves of the LCO, LCO-A5, LCO-A10, and LCO-A20 at the first, 20th, 40th, 60th, 80th, and 100th cycles at 2.5 μA cm–2 are shown in Figure a, b and Figure S2a, b. The first-cycle discharge specific capacities of the LCO, LCO-A5, LCO-A10, and LCO-A20 are 46.6, 44.3, 45.7, and 43.3 μAh μm–1 cm–2, respectively. At the same time, it can be observed that the voltage of the discharge platform of the film after proper doping (such as in the LCO-A5 and LCO-A10 samples) is more stable, and the capacity attenuation is slower. Figure c shows the 110-cycle cycle performance and 110-cycle capacity retention rate of the LCO, LCO-A5, LCO-A10, and LCO-A20 under the same conditions. It can be observed that the cycling performance of the LCO-A5 and LCO-A10 samples was significantly improved with capacity retention of 94.58 and 94.96% after 110 cycles, respectively, compared to 80.04% for the LCO, which is mainly due to the partial substitution of Co sites by doped Al to prevent their dissolution from the LiCoO2 lattice.[39] The formed LiCo1–AlO2 solid solution stabilizes the structure of the film during cycling. In addition, the doping amount of the LCO-A20 sample is too large, resulting in poor crystallinity of the LiCoO2 film, and thus its cycling performance is poor.

Figure 4

Discharge curves of (a) LCO and (b) LCO-A10. (c) Cycling performance of doped and undoped electrodes.

To explore the effect of Al doping on the diffusion rate of lithium ions in LiCoO2 thin films, we subjected the undoped and doped modified samples to variable scan speed CV test analysis at 0.05, 0.1, 0.2, 0.4, and 0.6 mV s–1, the results are shown in Figure . It can be observed that with an increasing CV scan rate, the redox peak potential of the sample moves to the two sides, which reflects the change in polarization.[40]Figure e shows the relationship between the reduction peak current Ip and the square root of the scan rate (v1/2). All samples show a good linear relationship, which is a typical diffusion control process. The following formula can be used to calculate the lithium-ion diffusion coefficient DLi.[41]where Ip is the peak current (A), n is the number of transferred charges (n = 1), F is the Faraday constant (96485.4 C mol–1), S is the area of the electrode sheet (cm2), CLi is the lithium-ion concentration in the material (0.051 mol cm–3), R is the gas constant (8.314 J mol–1 K–1), T is the absolute temperature (K), DLi is the lithium-ion diffusion coefficient (cm2 s–1), and v is the scan rate (V s–1). The lithium-ion diffusion coefficients D for the LCO, LCO-A5, LCO-A10, and LCO-A20 samples were calculated using this formula to be 5.31 × 10–14 cm2 s–1, 1.21 × 10–13 cm2 s–1, 1.59 × 10–13 cm2 s–1, and 4.49 × 10–14 cm2 s–1, respectively. The lithium-ion diffusion coefficient of the film after proper doping is significantly improved because the Al doping makes the c-axis larger, and the lithium-ion diffusion coefficient increases significantly. The LiCo1–AlO2 solid solution formed stabilizes the LiCoO2 structure and provides a better diffusion pathway for lithium ions.[42] Due to a large amount of Al doping in the LCO-A20 sample, the structure of LiCoO2 is destroyed so that the diffusion rate of lithium ions does not increase but decreases. Due to a large amount of Al doping in the LCO-A20 sample, the LiCoO2 is less crystalline, so the diffusivity of lithium ions is not increased, but decreased.

Figure 5

Cyclic voltammetry profiles of (a) LCO, (b) LCO-A5, (c) LCO-A10, and (d) LCO-A20 at various scanning rates from 0.05 to 0.6 mV s–1. (e) Peak current Ip versus the square root of the scan rate (v1/2) the relationship diagram

To further investigate the effect of Al doping on the electrical conductivity of the LCO films, the electronic conductivities of the LCO, LCO-A5, LCO-A10, and LCO-A20 samples were tested by the four-probe method. Figure a shows the relationship between Al doping and electronic conductivity. The electronic conductivities of the LCO, LCO-A5, LCO-A10, and LCO-A20 samples were determined to be 8.9 × 10–5 s/cm, 1.096 × 10–4 s/cm, 1.29 × 10–4 s/cm and 2.76 × 10–4 s/cm, respectively, which clearly shows that the electronic conductivity of the LiCoO2 composite films increases with increasing Al content. The Al is a good conductor of electrons, and as the sputtering power of the Al target increases, the increase in Al content in the film tends to form a conductive network, thus improving the electrical conductivity of the film. However, in combination with the previous XRD patterns, it can be observed that the excessive Al content has a detrimental impact on the lattice structure of LiCoO2, resulting in poor cycling performance. Therefore, although the electronic conductivity is excellent, the effect of Al doping on the crystallinity and composition of the film needs to be considered.

Figure 6

(a) Relationship between electronic conductivity and Al content atomic percent of Al-doped LiCoO2 thin film, (b) EIS Nyquist plots of the LCO, LCO-A5, LCO-A10, and LCO-A20 before cycling

To further investigate the contribution of Al to the improved electrochemical properties of the LiCoO2 films, we performed precycle AC impedance spectroscopy tests for the samples before and after doping, and the results are shown in Figure b. The EIS spectra of the samples consist of a semicircle followed by a line portion. The tangent line between the semicircle and abscissa in the high-frequency band reflects the magnitude of the charge transfer impedance (Rct), and the oblique line in the low-frequency band represents the Warburg impedance.[43] The Rct impedance of the LCO-A5 and LCO-A10 electrodes is significantly lower than that of the undoped modified samples. The LiCo1–AlO2 solid solution formed by Al doping can stabilize the change in the LiCoO2 structure during the process of lithium-ion deintercalation and is beneficial to the diffusion of lithium ions. On the other hand, the Al shows good electronic conductivity, and the Al that does not enter the LiCoO2 lattice in the composite film prepared by cosputtering forms a conductive channel in the film, which improves the electronic conductivity of the composite film and reduces its impedance. Combined with previous XRD and cycle performance diagrams, it can be observed that the Al content of the LCO-A20 sample is relatively large, which destroys the structure of the LiCoO2 film and hinders the migration of lithium ions, so the impedance is large.

Al2O3 In Situ Coating of Al-Doped LiCoO2 Thin Films

To further improve the surface interface stability of the films, we coated the Al-doped LCO films with Al2O3 in situ. A Raman test was first used to probe the physical phase structure and surface composition of the Al2O3-coated LCO-A10 electrode, and the results are shown in Figure S3. The two characteristic peaks observed in the figure are typical of the HT-LCO phase. The Eg peak located near 487 cm–1 indicates the bending vibration mode of Co–O, and the A1g peak located near 597 cm–1 indicates the stretching vibration mode of O–Co–O.[44] This shows that the LiCoO2 thin films have a high-temperature phase layered structure and that an appropriate amount of Al cosputtering doping and the Al2O3 coating has no obvious effect on the crystal structure of LCO. An additional small peak observed at approximately 1180 cm–1 indicates the presence of a trace amount of Co3O4 impurity phase, which may be due to the loss of lithium during magnetron sputtering.[26]

To demonstrate the presence of the Al2O3 coating layer on the film surface, we performed the XPS tests on the LCO-A10@Al2O3 and LCO samples, and the results are shown in Figure . From the Li 1s spectra and Co 2p spectra shown in Figures a, b, it can be observed that both the LCO-A10@Al2O3 and LCO samples show characteristic peaks associated with Li+ near 54 eV and Co3+ near 780.5 and 795.5 eV.[45] A comparison shows that the binding energy of Li+ and Co3+ does not change significantly after Al doping and in situ coating, but the presence of the Al2O3 coating layer blocks the detection signal, resulting in a weaker peak intensity for the relevant characteristic peaks in the Li 1s and Co 2p spectra for the LCO-A10@Al2O3 sample compared to that for the LCO sample. In the O 1s spectra of Figure c, the characteristic lattice oxygen peak of LCO is around 529.7 eV. The LCO-A10@Al2O3 sample forms a LiAlCo1–O2 solid solution due to the Al doping, which makes the Co–O bond more covalent and causes the peak of lattice oxygen to shift toward the high binding energy.[22] A peak on the high binding energy side of the O 1s spectrum of the LCO sample is found, which can be attributed to O–H bonds from hydroxyl groups or water molecules adsorbed on the electrode surface. In addition, the LCO-A10@Al2O3 sample 532.3 eV peak is the characteristic peak of Al2O3, indicating that we have successfully deposited the Al2O3 cover layer on the surface of the film by reacting the sputter.[46,47] As shown by the Al 2p spectra in Figure d, the LCO sample does not show Al-related feature peaks. However, the characteristic peak of LiAlCo1–O2 solid solution appeared in LCO-A10@Al2O3 near 73.4 eV, and the characteristic peak of Al2O3 appeared near 74.3 eV.[45] The successful doping of Al into the LCO lattice to form a LiAlCo1–O2 solid solution is not only demonstrated, but the successful coating of the LCO-A10@Al2O3 electrode with an Al2O3 coating layer is once again demonstrated.

Figure 7

(a) Li 1s, (b) Co 2p, (c) O 1s, and (d) Al 2p XPS spectra of LCO and LCO-A10@Al2O3.

Figure a shows the 240-cycle performance of the LCO, LCO-A10, and LCO-A10@Al2O3 samples measured at 2.5 μA cm–2. The first-cycle discharge specific capacities of the LCO, LCO-A10 and LCO-A10@Al2O3 electrodes are 46.6, 45.7, and 42.7 μAh um–1 cm–2, respectively. The discharge capacity decreases after coating with Al2O3 because the existence of the coating layer leads to an activation process in the electrode discharge process. From the cycle performance of the LCO-A10@Al2O3 electrode shown in Figure a, it can also be observed that the electrode exists in the rising activation state for the first 10 cycles. The 110-cycle capacity retention rate of the LCO is only 80.04%, while the 240-cycle capacity retention rate for the LCO-A10 and LCO-A10@Al2O3 are 81.83% and 94.14%, respectively. This result is superior to the literature report.[25,48] The cycling stability of the LCO-A10@Al2O3 is significantly improved compared to both the LCO and LCO-A10, indicating that after stabilizing the LiCoO2 structure and improving its conductivity by Al doping, the electrode surface interface is stabilized by Al2O3 coating modification, which is more conducive to improving the electrode cycling performance. Figure b shows a graph of the rate performance of the LCO, LCO-A10, and LCO-A10@Al2O3 samples. It can be observed that LCO-A10 exhibits the best rate performance with a discharge specific capacity of 43.5 μAh um–1 cm–2 at 100 μA cm–2. This is because Al doping improves both the electronic and lithium-ion conductivities of the LCO films, indicating that our cosputtering doping strategy can not only effectively improve the cycle performance of the LCO films but also significantly improve its rate capability. The rate performance of the LCO-A10@Al2O3 electrode is decreased compared with that of the LCO-A10 electrode because although Al2O3 has good inertness in the electrolyte, which can reduce the erosion of electrolyte to the electrode, at the same time it is poor lithium-ion and electron conductivity has an adverse influence on lithium-ion and electron diffusion, which leads to the decrease in its high rate performance. However, the rate performance of LCO-A10@Al2O3 is still significantly improved compared to that of the LCO electrode.

Figure 8

Electrochemical performance of the LCO, LCO-A10, and LCO-A10@Al2O3: (a) 240-cycle performance at 2.5 μA cm–2, (b) rate performance

The charge–discharge midpoint potentials of LCO and LCO-A10@Al2O3 are shown in Figure . The charging midpoint potential shows an increasing trend due to the polarization during charging and discharging.[49] The discharge midpoint potential is a common index to evaluate the potential decay,[50] and the discharge midpoint potential of LCO decreases sharply from 3.9314 to 3.8596 V after only 110 cycles, ΔV1 = 0.0718 V, the discharge midpoint potential of LCO-A10@Al2O3 slowly decreases from 3.8985 to 3.8611 V (ΔV2 = 0.0374 V) after 240 cycles. During cycling, the midpoint potential difference of LCO drastically increases to 0.1424 V after only 110 cycles, whereas those of LCO-A10@Al2O3 only gently grow to 0.1316 V after 240 cycles. This indicates that the LCO-A10@Al2O3 electrode exhibits good voltage stability during charge–discharge cycles, which may be due to the fact that the Al2O3 coating stabilizes the electrode surface and improves the structural stability of the material.

Figure 9

Midpoint potentials of LCO and LCO-A10@Al2O3 at 2.5 μA cm–2.

To study the effect of the Al2O3 coating on the diffusion rate of lithium ions, we carried out a variable-sweep rate CV test for the samples after in situ coating, and the test results are shown in Figure a. Figure b shows a graph showing the relationship between the reduction peak current value Ip and the square root of the scan rate (v1/2). LCO-A10@Al2O3 sample shows a good linear relationship. The lithium-ion diffusion coefficient for the LCO-A10@Al2O3 sample is calculated using formula to be 9.18 × 10–14 cm2 s–1, which is lower than that of the LCO-A10 sample (1.59 × 10–13 cm2 s–1), but the lithium-ion diffusion coefficient is still larger than that of the unmodified sample (5.31 × 10–14 cm2 s–1). The above results show that the Al2O3 in situ coating can effectively stabilize the surface and interface of the LCO electrodes but also affects the migration of lithium ions to a certain extent. This also explains why the cycle performance of the LCO-A10@Al2O3 sample is significantly improved relative to that of the LCO-A10 sample, but the rate performance is decreased.

Figure 10

(a) CV curves measured for the LCO-A10@Al2O3 electrode with variable scan speeds, (b) LCO-A10@Al2O3 the relationship between the peak current Ip and the square root of the scan speed (v1/2).

Electrochemical impedance spectroscopy (EIS) of the LCO, LCO-A10 and LCO-A10@Al2O3 electrodes were tested and the results are shown in Figure . Figure a shows the impedance spectrum before cycling, and Table S1 gives the fitted value. Since Al2O3 is not a good Li-ion conductor, the impedance of the sample after coating increases compared to that of the LCO-A10 sample due to the presence of the coating layer. Figure b, c shows the impedance diagrams for the 20th and 50th, respectively, and Table S2 gives the impedance fitting value. The results show that the increase in the impedance of the LCO-A10 electrode is significantly smaller than that of the LCO electrode, regardless of Rf or Rct, which is due to the stabilization of the LiCoO2 structure by the Al doping. On the other hand, compared with the LCO-A10 electrode, the LCO-A10@Al2O3 electrode shows a smaller increase in cycle resistance. Because of the protection of the Al2O3 coating, the Rf value for the LCO-A10 electrode after cycling is significantly lower than that for the LCO and LCO-A10 electrodes, with a value of only 21.1 Ω obtained after 50 cycles, while the Rf values of the LCO and LCO-A10 electrodes show an increase to 2551.0 and 1125.2 Ω, respectively. The results show that the side reaction on the surface of the LCO and LCO-A10 electrodes is serious, and an unstable CEI film is formed, which greatly increases the impedance of the surface layer, while the existence of the Al2O3 coating stabilizes the electrode surface interface. Impedance tests show that Al doping can stabilize the crystal structure of the LiCoO2 thin film and reduce the increase in impedance. Al2O3 coating of the Al-doped electrode can further reduce the side reactions on the electrode surface, stabilize the electrode surface, and further reduce the increase in impedance.

Figure 11

EIS Nyquist plots of the LCO, LCO-A10, and LCO-A10@Al2O3 thin-film electrodes: (a) before cycling, (b) 20th, and (c) 50th.

Conclusions

In summary, we successfully prepared Al-doped and in situ Al2O3 coated LiCoO2 composite thin-film electrodes by magnetron sputtering technology, which significantly improved its electrochemical performance. Compared with an unmodified LCO film, the Al cosputtered doped LiCoO2 film showed better cycle performance and rate performance, among which the LCO-A10 electrode showed a 110-cycle capacity retention rate of 94.96% (the capacity retention rate of the LCO is 80.04%). Al doping improves the electronic and lithium-ion conductivity of LiCoO2 films, and the formation of the LiAlCo1–O2 solid solution by modification of LiCoO2 films is beneficial in improving the stability of the film structure. Furthermore, the interface of doped LiCoO2 film was further stabilized by the in situ coating of Al2O3. The double-modified composite film electrode showed a capacity retention rate of 94.14% (40.62 μAh um–1 cm–2) after 240 cycles at 2.5 μA cm–2. The in situ continuous deposition of the Al2O3 thin film coating layer uniformly transitions at the interface between the electrode and the coating layer and does not form an obvious physical and chemical contact interface, which is conducive to the stability of the coating interface, inhibits the increase in the resistance of the electrode during cycling, and further improves the cycle stability of LiCoO2. The results show that doping and in situ coating of LiCoO2 thin films by magnetron sputtering is an effective modification method, which provides useful inspiration for the development of high-performance thin-film electrode materials.