Air stable Al 2 O 3 -coated Li 2 NiO 2 cathode additive as a surplus current consumer in a Li-ion cell

Transcription

1 PAPER Journal of Materials Chemistry Air stable Al 2 O 3 -coated Li 2 NiO 2 cathode additive as a surplus current consumer in a Li-ion cell Min Gyu Kim a and Jaephil Cho* b Received 14th August 2008, Accepted 30th September 2008 First published as an Advance Article on the web 5th November 2008 DOI: /b814161d Highly air stable Al 2 O 3 -coated Li 2 NiO 2 cathode additive is prepared by coating with Al iso-propoxide on Li 2 NiO 2, obtained from firing of a physical mixture of pure Li 2 O and NiO at 600 C for 10h under N 2 atmosphere. An as-prepared coated cathode has first charge and discharge capacities of 420 mah/g and 310 mah/g, respectively, between 4.3V and 1.5V showing an irreversible capacity ratio of 26%. However, when the discharge cut-off voltage increases to 2.75V (2.85V vs. graphite), its discharge capacity decreases to 120 mah/g, which corresponds to an irreversible capacity ratio of 71%. Owing to such a high irreversible capacity, it can effectively compensate for the irreversible capacity of the Li-ion cell using LiCoO 2 and natural graphite as cathode and anode materials, respectively, in spite of only 4wt% addition to the LiCoO 2 cathode. In addition, the additive prevents the 12V overcharge, thereby preventing the explosion of the cell. We believe that Li 2 NiO 2 decomposition consumes the surplus current during the overcharging to 12V, and therefore the voltage is not increased until the complete decomposition of the Li 2 NiO Introduction Most Li secondary batteries use LiCoO 2 as a cathode material due to a high volumetric energy density and excellent hightemperature performance in contrast to LiNi 1-x M x O 2 and LiMn 2 O 4, for example, such as the cycle life at 60 C and the swelling characteristics at 90 C. 1 3 However, the safety of Li secondary batteries is still a major concern, particularly as the capacity of such batteries increases. A major problem occurring during Li-ion cell operation concerns the abrupt overcharging of the voltage supply limit (12 V) due to a defect or malfunction in the protective devices of the cell. 4 Such disasters occur primarily because LiCoO 2 cathodes can undergo a violent exothermic reaction with an electrolyte during the overcharging process, which may result in a short-circuiting of the cell. 5 9 A combination of the temperature increase and the internal short circuit of the cell eventually results in an explosion of the cell. We previously suggested that an AlPO 4 nanoparticle coating on LiCoO 2 or LiNi 0.8 Co 0.1 Mn 0.1 O 2 could reduce the 12 V overcharge instability of Li-ion cells In addition, many studies have endeavored to improve the thermal stability of Li-ion cells by adding electrolyte additives, such as 3-chloroanisole (3CA), biphenyl (BP) and cyclohexyl benzene (CHB) However, both the coating and the electrolyte additives led to processing cost increase and degraded electrochemical properties. Recently, physical blending of LiCoO 2 with an orthorhombic Immm Li 2 NiO 2 was reported to suppress overdischarge below 1.5 V in a Li-ion cell. 20 This is only one study dealing with improving the safety of Li-ion cells using a cathode additive. a Beamline Research Division, Pohang Accelerator Laboratory, Pohang, Korea b Department of Applied Chemistry, Hanyang University, Ansan, Korea jpcho@hanyang.ac.kr More importantly, one of the fundamental problems of Ni-rich phases is the rapid reaction with air resulting in the formation of Li 2 CO 3 and LiOH on the surface It has been reported that surface active oxygen O 2 from an impurity NiO phase in LiNiO 2 2 combined with CO 2 and H 2 O in air to form CO 3 and OH, and suggested the following surface reaction mechanism: 2Li CO 3 /2OH / Li 2 CO 3 /2LiOH. 24 The presence of such impurities led to severe cell swelling during the formation process in the Li-ion cell manufacturing and at >60 C storage at the charged state. In this regard, Li 2 NiO 2 is expected to have more severe reactivity with moisture than any other Ni-based cathodes. In the present study, coating Al 2 O 3 onto cathode additive Li 2 NiO 2 shows significantly decreased moisture reactivity upon air exposure. Moreover, the physical addition of 4 wt% Al 2 O 3 - coated Li 2 NiO 2 additive into a LiCoO 2 cathode can achieve 100% coulombic efficiency without any explosion during 12 V overcharge test of Li-ion cells, while a cell without Li 2 NiO 2 additive explodes at 12 V overcharge. 2. Experimental The electrochemical behavior of Li 2 NiO 2 is highly dependent on the purity of the Li source, such as Li 2 O, which is essential to obtain a high capacity of Li 2 NiO 2. In the synthesis process, therefore, a careful approach to pre-treatment of the Li precursor is important since any content of other Li phases may downgrade the electrochemical properties. Unlike a previous study which used a commercially available Li 2 O, the Li 2 O with high purity of 99.9% in the present study was obtained from the thermal decomposition of Li 2 CO 3 (99.99%) at 700 C in an oxygen atmosphere. All commercially available Li 2 O showed a detectable level of LiOH impurity. For instance, an orthorhombic Immm Li 2 NiO 2 prepared from a mixture of Li 2 O (Aldrich, 97%) 5880 J. Mater. Chem., 2008, 18, This journal is ª The Royal Society of Chemistry 2008

2 and NiO (Aldrich, 99%) showed that the charge and discharge capacities at the first cycle are about 370 and 120 mah/g, respectively between 4.3 and 3V (no cycling data were provided). 20 However, the cathode prepared by the method according to ref. 20 showed a large amount of unreacted NiO phase, and the capacities were 250 mah/g and 60 mah/g, respectively. For preparing uncoated Li 2 NiO 2, 37g of NiO (average particle size was 10 mm) and 15g of as-synthesized Li 2 O was thoroughly mixed by using an automatic mixing machine in an N 2 stream. The mixture then was fired at 600 C for 10h under pure N 2 atmosphere. For preparing the coated sample, 1.5g of Al iso-propoxide was dissolved in 50g of ethanol. 50g of Li 2 NiO 2 was then put into the solution which was mixed for 20 min, followed by drying at 120 C for 12h. The rest of the synthesis method is identical to that of the uncoated Li 2 NiO 2. Liquid electrolytes (1M LiPF 6 in ethylene carbonate (EC)/ dimethyl carbonate (DMC)/diethyl carbonate (DEC) (1/1/1, w/w)) (Cheil Industries, Korea) containing less than 20 ppm HF were used for electrochemical tests. The cathodes for the battery test cells ( size prismatic cell) were made from the LiCoO 2 cathode material and a mixture of LiCoO 2 (average particle size was 18 mm) and Al 2 O 3 -coated Li 2 NiO 2 with a weight ratio of 96:4, Super P carbon black, and polyvinylidene fluoride (PVdF) binder (Kureha Company) in a weight ratio of 96:2:2. The anodes for the battery test cells were made from natural graphite powders and polyvinylidene fluoride (PVdF) binder in a weight ratio of 94:6. The dimensional ratio of anode to cathode of 1.08:1 was used for the Li-ion cells with and without the cathode additive. The charge capacities of the Li-ion cells were set at 940 mah. The electrodes were prepared by coating a cathode slurry onto an Al foil followed by drying at 130 C for 20 min. The slurry was prepared by thoroughly mixing an N-methyl- 2-pyrrolidone (NMP) solution of PVdF, carbon black, and the powdery cathode material. The test cells were aged at room temperature for 24 h prior to the electrochemical test. Li-ion cells with a nominal capacity of 870 mah or 945 mah were used for the electrochemical tests and 12 V overcharging experiments. The overcharge test was conducted with a constant current of 1 A. The cell-surface temperature was monitored using a K-type thermocouple placed on the center of the largest face in the cell can, and the thermocouple was tightly glued with insulating tape. For the rate capability and life cycle performance tests, the cells were charged with a constant current mode to 4.2 V and maintained at this voltage with a constant voltage mode (CV) for 5h. Inductively coupled plasma-mass spectroscopy (ICP, ICPS- 1000IV, Shimadzu) was used to determine the Al content in the coated Li 2 NiO 2, and the coated sample usually contained 1wt% Al 2 O 3 (99 wt% Li 2 NiO 2 ). Electrochemical in situ cell preparation for XAFS measurement was carried out in an inert-gas filled glovebox to prevent any oxidation and contamination. After assembly with Li foil, separator and Al 2 O 3 -coated Li 2 NiO 2 electrode, the cell was in-vacuum sealed with an aluminium pouch. X-Ray photon flux was enough to transmit through the electrochemical cell without any interaction. Ni K-edge X-ray absorption spectra were recorded on the BL7C1 beam line of Pohang light source (PLS) with a ring current of ma at 2.5 GeV. Si(111) double crystal monochromator was employed to monochromatize the X-ray photon energy. The data were collected in transmission mode with N 2 gas-filled ionization chambers as detectors. Higher order harmonic contaminations were eliminated by detuning to reduce the incident X-ray intensity by 30%. Energy calibration was simultaneously carried out for each measurement with reference Ni metal foil placed in front of the third ion chamber. The data reduction of the experimental spectra was performed by the standard procedure reported previously. The moisture (OH ) in the sample was determined using a Karl Fisher moisture titrator at 250 C; prior to measuring, the sample was vacuum-dried at 150 C for 2 h to remove the H 2 O molecules adsorbed on the sample. Therefore, the measured moisture was from OH in LiOH. 3. Results and discussion Fig. 1 shows XRD patterns of the as-synthesized and Al 2 O 3 - coated Li 2 NiO 2 powders, and the overall patterns can be indexed to an orthorhombic phase form with a space group of Immm although small peaks for NiO phase were observed. 25 The lattice constants before and after coating are quite similar, with values of a ¼ Å, b ¼ Å, and c ¼ Å. Fig. 2 shows SEM images of the as-synthesized (a, b, c) and Al 2 O 3 -coated Li 2 NiO 2 powders (d, e, f). The SEM images of the bare samples show aggregated caterpillar-liked particles with a length of 0.5 mm and a diameter of 80 nm. On the other hand, the caterpillarliked particles after the Al 2 O 3 coating appear to disappear (e) and the surface of sample becomes smooth. An expanded image of e (f) shows the caterpillar-liked particles, but the surface morphology is obviously different from the uncoated one, indicating that the Al 2 O 3 coating layer had reacted with the bulk Li 2 NiO 2, thus forming a solid solution such as Li O Al Ni O. Fig. 3 shows the voltage profiles of the bare and Al 2 O 3 -coated Li 2 NiO 2 between 1.5 V and 4.3 V during the first cycle. The bare Li 2 NiO 2 cathode in the present study has a charge capacity of 337 mah/g and a discharge capacity of 250 mah/g which are similar to the capacity values reported earlier. 20,25 However, the Al 2 O 3 -coated cathode shows increased initial capacities with respect to those of bare Li 2 NiO 2, exhibiting a charge capacity of 420 mah/g and a discharge capacity of 310 mah/g. However, when the discharge cut-off voltage increases to 2.75V (2.85V vs. Fig. 1 XRD patterns of the as-synthesized bare and Al 2 O 3 -coated Li 2 NiO 2 powders. This journal is ª The Royal Society of Chemistry 2008 J. Mater. Chem., 2008, 18,

3 Fig. 2 SEM images of the as-synthesized (a, b, c) and Al 2 O 3 -coated Li 2 NiO 2 powders (d, e, f). cycling between 4.3 V and 1V. Hence, it may be plausible that the origin of the two plateaus is not from such phase transitions. The higher deintercalated amount of the coated sample than the uncoated one is related to the coating layer which stabilizes the structure under electrolyte. However, a problem for scale-up with the uncoated cathode was rapid moisture uptake upon exposure to air. After 7h exposure to air, the moisture of the uncoated sample rapidly increases from 260 ppm to ppm, values which are much higher than those for LiNi 1-x M x O 2 cathodes. 26,27 However, its amount in the coated cathode only rises from 250 ppm to 300 ppm after same time exposure as the uncoated sample. The FT-IR spectrum of the coated sample showed no changes even after exposure to air for 7h, but that of the bare sample showed significantly increased peaks of water and CO 2 at 3500 cm 1 and 2500 cm 1, respectively (Fig. 4). The moisture uptake of the cathode powders leads to the formation of non-conducting Li 2 CO 3 and LiOH phases on the particle surfaces, resulting in a decrease in capacity. 26,27 The uncoated sample exposed to air for 7h showed a rapid capacity decrease to Fig. 3 (a) Voltage profiles of bare and Al 2 O 3 -coated Li 2 NiO 2 during the first cycle at a rate of 0.2 C in coin-type half cells and (b) voltage profiles of the coated cathodes between 4.3 and 1.5 V at a rate of 0.2 C in a cointype half cell after 1, 3, 5, 10 cycles. graphite), the discharge capacity decreases to 120 mah/g, which corresponds to an irreversible capacity ratio of 71%. In addition, it showed quite reversible capacity to 10 cycles (Fig. 3b). Expect for the first charge, plateaus at 4 and 3V regions were reported to be from phase transformations from layered LiNiO 2 to layered NiO 2 (about 3.9 V) and from layered LiNiO 2 to Immm-Li 2 NiO 2 (experimentally about 1.9 V). In spite of the distinct two step regions, capacity fade to 10 cycles is excellent, and discharge capacities after the first and 10th cycles are 290 and 281 mah/g, respectively. In contrast, LiMn 2 O 4 which has a cubic to tetragonal phase transition at 2.8V showed rapid capacity fade when Fig. 4 FT-IR spectra of the bare and coated cathode powders after synthesis (moisture was measured after completely cooling the sample to room temperature in the furnace) and after 7h exposure to air (relative humidity was 40%) J. Mater. Chem., 2008, 18, This journal is ª The Royal Society of Chemistry 2008

4 Fig. 5 Coulombic efficiencies of natural graphite and LiCoO 2 in cointype half cells for the first cycle. LiCoO 2 shows the voltage profile during charge to 4.3V. The cells were cycled at a rate of 0.2C. the charge capacity of 120 mah/g, while the coated one showed no capacity decrease. The cathode Li 2 NiO 2 additive can be expected to compensate for the inherent capacity loss from the anode material in a Li-ion cell, by donating enough Li-ion through the sacrifice of its irreversible capacity. Fig. 5 shows the typical voltage profiles of natural graphite and LiCoO 2 in coin-type half cells for the first cycle using Li metal as counter electrode. In general, it is wellknown that the first discharge capacity of pure LiCoO 2 inherently decreases to about 91% in the Li-ion full cell with assembly of anode graphite and cathode LiCoO 2 (160 mah/g 0.91 ¼ 146 mah/g), due to the irreversible capacity ratio of natural graphite as a counterpart electrode. It is because the Li-ion intercalated into layered graphite in the initial charging state (in the viewpoint of cathode) cannot be perfectly re-donated to delithiated Li 1-x CoO 2 in the following discharge state, leading to an irreversible capacity of about 9%. In order to compensate the inevitable initial capacity loss in the graphite pure LiCoO 2 full cell assembly, the above Al 2 O 3 -coated Li 2 NiO 2 with its high irreversible capacity in the first cycle can be effectively used as a cathode additive. It is also anticipated that the Al 2 O 3 -coated additive donates higher capacity to natural graphite than the bare material without additive, leading to reduction of the irreversible capacity of the graphite anode. Note that the irreversible capacity ratio is 71% between 4.3 V and 2.85V (between 4.2 V and 2.75V vs. graphite) in Fig. 3a. In order to see the effect of the cathode additive in a real Li-ion full cell, the charge capacities with and without the cathode additive were set at 945 mah in prismatic Li-ion cells. Fig. 6a and b show voltage profiles of the balanced Li-ion cells with only LiCoO 2 and a mixture of LiCoO 2 and Al 2 O 3 -coated Li 2 NiO 2 (with a weight ratio is 96:4) between 2.75 V and 4.2 V for the formation cycle. (All the assembled fresh cells were aged for 2 days at 21 C and cycled at a rate of 0.2C per cycle.) As expected, the Li-ion cell without the additive has a coulombic efficiency of 92%, a value similar to that observed in a coin-type half cell (Fig. 3a). However, when 4 wt% of Li 2 NiO 2 was added to the LiCoO 2, we observed no irreversible capacity ratio, showing the same charge and discharge capacities of 945 mah (Fig. 6b). This means that the charge capacity from the Li 2 NiO 2 is fully donated to the anode, thereby compensating the irreversible capacity loss of the anode. Even with the addition of the coated Li 2 NiO 2,we Fig. 6 Voltage profiles of the balanced Li-ion cells with only (a) LiCoO 2 and (b) a mixture of LiCoO 2 and coated Li 2 NiO 2 (with a weight ratio is 96:4) between 2.75 V and 4.2 V for the formation cycle (all the assembled fresh cells were aged for 2 days at 21 C and cycled at a rate of 0.2C per cycle). (c) Rate capabilities and (d) cycle life performance of the Li-ion cell with the coated Li 2 NiO 2. For the cycle life performance, the cell was cycled at a rate of 1C (¼ 945 mah). found that the rate capability of the cell is not influenced by the additive and that the capacity retention at 2C is 96% (Fig. 6c). The cycle life performance of the Li-ion cell with the Al 2 O 3 - coated Li 2 NiO 2 exhibits a capacity retention of 87% (from 934 mah to 810 mah) after 250 cycles (Fig. 6d), which is comparable to the cell with LiCoO 2 only. 28 These results are similar to the results of Li-ion cells without Li 2 NiO Note also that the cell with the additive had an electrode density of 3.6 g/cc, which is similar to the electrode density of the cell without the additive. As a result, the cathode additive can positively contribute to overcoming the initial capacity loss without affecting the rate capability and electrode density. Fig. 7 shows ex situ XRD patterns of the Al 2 O 3 -coated Li 2 NiO 2 electrodes obtained at certain designated cut-off voltages during two cycles. At 3.9V, peaks intensities at 19.8 and 25.5 decrease abruptly and the peak intensity at 19 grows fast at 4V. After that the peak at 19 maintains its dominant intensity out to the 2 nd cycles, but the peaks are very broad. Since there was large irreversible capacity after 1 cycle, this broad peak is suspected to be responsible for that. In addition, the peaks of Li 2 NiO 2 becomes broad and overall peak intensities of the Li 2 NiO 2 phase also decrease, indicating a breaking-up of the long range order of the layered orthorhombic phase. In spite of charging to 4.3V, the Li 2 NiO 2 phase is still observed indicating This journal is ª The Royal Society of Chemistry 2008 J. Mater. Chem., 2008, 18,

5 Fig. 7 Ex situ XRD patterns of the Al 2 O 3 -coated Li 2 NiO 2 electrodes obtained at some selected cutoff voltages during two cycles. no phase transformation. Also, note the increased peak intensity of the NiO upon cycling. This means the bulk structural variation has not been changed in the Li-ion behavior. Fig. 8 shows ex situ XRD patterns after 5 and 10 cycles, in which the Li 2 NiO 2 phase is observed in contrast to the result of bare Li 2 NiO 2 reported earlier. 21 They suggest that the Li 2 NiO 2 phase except for the impurity NiO phase was transformed into Fig. 8 XRD patterns of the coated Li 2 NiO 2 electrodes after 5 and 10 cycles. a material with very small crystallite size or with a fully amorphous structure after the first cycle. However, the present study of Al 2 O 3 -coated Li 2 NiO 2 clearly shows the existence of Li 2 NiO 2 even after 5 and 10 cycles. On the other hand, upon further cycling, the Li 2 O phase is observed, which is not observed up to 2 cycles, indicating the possible decomposition of the Li 2 NiO 2 phase upon cycling, which may be attributed to capacity fade. In order to investigate local structural variation of highly irreversible cathode additive Li 2 NiO 2 during Li reactions, in-situ Ni K-edge XAFS experiments were carried out. The voltage profile as a function of electrochemical reaction time during the XAFS data collection is similar to the original electrochemical property. Fig. 9 shows spectral variations of the X-ray absorption near edge structure (XANES) and the Fourier transform magnitudes (FT) of k 3 -weighted extended X-ray absorption fine structure (EXAFS) for cutoff voltage-resolved Ni K-edge XAFS spectra. For the pristine Al 2 O 3 -coated Li 2 NiO 2, the peak features A, B, C (C 1 and C 2 ) are typically characteristic of the square planar symmetry, D 4h, around the central Ni 2+ ion. Based on earlier reports for Ni and Cu complexes with D 4h local symmetry, all the peaks can be reasonably assigned as follows The weak absorption peak A around 8333eV represents the transition of a 1s electron to a 3d hole state in 3d 8 electronic configuration for Ni 2+ ion, corresponding to 3d x2-y2 in low spin (S ¼ 0) and both 3d x2-y2 and 3d xy in high spin (S ¼ 1). The peaks B and C appear by the electric dipole-allowed transition of a 1s core electron to an unoccupied 4p bound state. The peak B at 8339eV corresponds to the final state of a 1s 1 c3d 9 L4p 1 with shakedown process by ligand to metal charge transfer (LMCT), where c and L represent a 1s core hole and a oxygen 2p ligand hole. The strong peak B occurring as a shoulder in the lower energy region with respect to main peak C (open circles in Fig. 9a), is a diagnostic peak feature for the square planar D 4h symmetry. As shown in Fig. 9a, the XANES features in the charging cycle have been changed effectively with respect to those of the pristine. The peak A is gradually shifted toward higher energy region, and the intensity of peak B, constant until 3.7V, begins to become weak and abruptly decreases upon charging to 3.9V. The peak features present the gradual increase of average oxidation state of Ni ion and before 3.9V the Li-ion can be simply extracted from the Immm structure maintaining square planar symmetry around the Ni ion and the local structure of square planar D 4h is effectively collapsed after higher delithiation to 3.9V. The peak features in the delithiated state are similar to those of edgeshared octahedral symmetry. Therefore it can be deduced that the delithiation leads to the flexibility of the square planar NiO 4 and the formation of octahedral symmetry by re-arrangement of square planar couples in the layered molecular sheets. Even in charging to 4.3V, however, peak B can be weakly observed, which means the residues of square planar symmetry exist in the highly delithiated state, corresponding to Li 0.36 NiO 2. Fig. 9b shows spectral comparison for the first cycling of the Al 2 O 3 - coated Li 2 NiO 2, in addition to those of reference materials. The following discharge to 2.0V does not show any dominant spectral variation and then any reversible return of spectral features to the pristine state, meaning the collapsed square planar symmetry can not be recovered in the consecutive Li-ion re-insertion. Therefore the XANES spectral features for the first cycle are 5884 J. Mater. Chem., 2008, 18, This journal is ª The Royal Society of Chemistry 2008

6 Fig. 9 Normalized Ni K-edge XANES spectra for spectral variations of the X-ray absorption near edge structure (XANES) for (a) the charge process (arrow in the lower inset figure indicates continuous increasing voltages until 4.3V during discharge, and the upper inset figure shows pre-edge peak features for the pristine and the charged state of 4.3V), (b) discharge process, and (c) the Fourier transform magnitudes (FT) of k 3 -weighted extended X-ray absorption fine structure (EXAFS) for selective cutoff voltage-resolved Ni K-edge XAFS spectra. closely related to the highly irreversible capacity of Li 2 NiO 2 itself. The geometric local structures around the Ni atom have been investigated with the Fourier-transformed (FT) magnitude of normalized Ni K-edge k 3 c(k) spectra, as shown in Fig. 9c. The FT peaks are selected for specific cutoff voltages. For pristine Al 2 O 3 -coated Li 2 NiO 2, the first three large FT peaks between 1.0 and 4.0 Å are important to determine the correct structural information since the FT peaks are closely indicative of the D 4h square planar site symmetry around the central Ni atom. The FT peak at 1.5Å corresponds to four-coordinated oxygen of the nearest neighboring atom around the Ni atom, while the FT peak at 2.4Å is assigned to the contribution of Ni Ni A by way of edge-shared coordination with two neighboring square planes. The FT peak at 3.3 Å means direct single scattering with Ni B atoms which are located in the neighboring plane. The higher FT peaks above 4.0Å include practically extended Ni Ni A (or Ni B ) single and multiple scatterings. Therefore, local structural variation of Al 2 O 3 -coated Li 2 NiO 2 during Li-ion reaction can be investigated with the diagnostic FT spectral variation of the square planar symmetry in the Immm phase. Until charged to 3.7V, the FT peak features are constant with respect to that of the pristine sample except for the overall decrease of FT peak intensity, which means the Li-ion is simply This journal is ª The Royal Society of Chemistry 2008 J. Mater. Chem., 2008, 18,

7 extracted from the Immm phase without structural change. Upon charging to 4.3V, however, the FT peak features are effectively changed. The FT peak of Ni Ni A is shifted towards higher bond distance corresponding to a single scattering between edge-shared octahedra. The FT peak intensity of Ni Ni B abruptly decreases and furthermore the FT peak of multiple scattering for Ni Ni A Ni A disappears. These facts suggest that the higher Li-ion extraction after 3.7V leads to the partial pulverization of the Immm phase and then new formation of edge-shared octahedra by rearrangement and orientation of free square planar units in the Ni layers. These spectral results can be supported with XANES of the disappearance of peak B. Upon discharging to 3.8V, overall FT peak feature resembles a NiO phase including the generation of a new FT peak at 1.8 Å (Ni 2+ O in NiO phase). Also the FT peak of Ni Ni B between the edge-shared square planes is weakly re-developed, in addition to the constant existence of Ni O bonding for Li 2 NiO 2 at 1.4 Å. This FT peak feature shows the Li-ion re-insertion leads to not only recovery of the short-range ordered square planar through the pulverized Immm phase, but also generation of an electrochemically inactive NiO phase. The NiO phase is gradually formed until further Li-ion insertion to 2.0V. As a result, it can be deduced that the pristine Al 2 O 3 -coated Li 2 NiO 2 with square planar symmetry experiences the partial pulverization of the Immm phase and formation of edge-shared octahedra during Li-ion extraction, and then phase transition to a NiO phase followed by partial recover to D 4h symmetry during Li-ion insertion. After the first cycle, there is coexistence of the pulverized Immm phase with Li-ion defects and NiO. In the present study for Al 2 O 3 -coated Li 2 NiO 2, no phase transition to a layered R-3m LiNiO 2 phase earlier reported for bare Li 2 NiO 2 could be observed. Owing to the structural instability of Li 2 NiO 2, the cathode additive has a huge irreversible capacity during the first cycle; hence, it may improve the thermal stability of the cell during the charging process. When the Al 2 O 3 -coated Li 2 NiO 2 was charged to 12 V, it was completely decomposed to a NiO phase (Fig. 10). If this decomposition process consumes the surplus current during the overcharging phase, the voltage is not increased until the complete decomposition of the Li 2 NiO 2. This effect is very Fig. 10 XRD pattern of the coated Li 2 NiO 2 electrode after 12V charge. Fig. 11 (a) Profiles of the cell voltage, current, and the cell surface temperature during the overcharging test of Li-ion cells with and without the Li 2 NiO 2 additive. (b) Li-ion cell morphologies after 12V overcharge experiment. positive for the cell because during the decomposition voltage, heat dissipates out of the cell faster than heat accumulates in the cell, thereby preventing any explosion of the cell. For instance, the electrolyte additive 3CA acts as a surplus current consumer. 32 Fig. 11a shows profiles of the cell voltage and the cell surface temperature during the overcharging test of Li-ion cells that contain electrolytes, with and without the Al 2 O 3 -coated Li 2 NiO 2 additive. A major feature of Fig. 11a is the steep voltage increase above 6 V (up to 12 V). The internal temperature can rise above the melting temperature of a polymer separator (T m ¼ 120 Cto 150 C), leading to separator shutdown and a rapid increase in cell resistance. Moreover, a further increase of the internal temperature can even melt the separator and the Li metal (T m ¼ 180 C) deposited on the anode, eventually leading to an internal short circuit. After rising to 12 V, the voltage of the cell without the additive plunges to 0 V with a maximum cell-surface temperature of approximately 300 C. We found that five of the tested cells were severely distorted and that their cell covers were lost as a result of an explosion (Fig. 11b). However, the cells with the Li 2 NiO 2 additive showed no increase of cell voltage over 5.3 V for a period of 80 min and the cell surface temperatures did not exceed 100 C. In contrast to the cell without the Li 2 NiO 2 additive, the other cells had a less abrupt voltage increase up to 5.3 V and the voltage was sustained below 5.3 V for about 35 min. This behavior most likely occurs because Li 2 NiO 2 decomposition consumes the surplus current and consequently inhibits the voltage increase until the full decomposition of Li 2 NiO 2 to NiO. Note that the cell surface 5886 J. Mater. Chem., 2008, 18, This journal is ª The Royal Society of Chemistry 2008

8 temperature during this decomposition does not exceed 50 C. This phenomenon indicates that during Li 2 NiO 2 decomposition, the heat dissipation rate is faster than the heat accumulation rate in the cell, leading to decrease of the cell surface temperature. Similar behavior is observed when the electrolyte additive 3CA acts as a surplus current consumer. 32 Coinciding with this behavior is the fact that the cell surface temperature reached its maximum of 100 C when the voltage reached 12 V. Furthermore, after the test, the cell morphology was identical to the morphology that existed before the charging, except for the swelling that occurred as a result of the gas evolution from the electrolyte decomposition (Fig. 11b). This result demonstrates that the Li 2 NiO 2 additive acts as a surplus current consumer and consequently increases the heat dissipation rate outside the cell. Conclusions Our results confirmed that an Al 2 O 3 coating on Li 2 NiO 2 can increase the charge capacity from 337 mah/g to 420 mah/g with only a small increase (20 mah/g) of discharge capacity. When we added 4 wt% of the cathode additive to LiCoO 2, the irreversible capacity of the Li-ion cell completely disappeared, and the rate capability and cycle life performance were similar to the cathode without the additive. In addition, the results of the 12 V overcharge test clearly showed that the cathode additive effectively consumed the surplus current in about 20 min, thereby preventing the cell from exploding at 12 V. The Al 2 O 3 -coated Li 2 NiO 2 can be also used to balance a cell with a high irreversible capacity of other types of cathode and anode materials. Acknowledgements This work was supported by the Korea Science and Engineering Foundation (KOSEF) grant funded by the Korea government (MEST) (NO. R ). We also acknowledge the Pohang Light Source (PLS) for the XAS measurement. References 1 J. Cho, Electrochem. Commun., 2003, 5, J. Cho, J. Power Sources, 2004, 126, J. Cho, Y. J. Kim, T.-J. Kim and B. Park, Angew. Chem. Int. Ed., 2001, 40, (a) Laptop Batteries Are Linked to Fire Risk, New York Times, March 15, 2001; U.S. Consumer Product Safety Commission ( (b) exploding+explosion. 5 R. A. Leising, M. J. Palazzo, E. S. Takeuchi and K. J. Takeuchi, J. Electrochem. Soc., 2001, 148, A H. Maleki, S. A. Hallaj, J. R. Selman, R. B. Dinwiddie and H. Wang, J. Electrochem. Soc., 1999, 146, A. Duaquier, F. Disma, T. Bowmer, A. S. Gozdz, G. G. Amatucci and J. -M. Tarascon, J. Electrochem. Soc., 1998, 145, H. Maleki, G. Deng, A. Anani and J. Howard, J. Electrochem. Soc., 1999, 146, J. R. Dahn, E. W. Fuller, M. Obrovac and U. Von Sacken, Solid State Ionics, 1994, 69, J. Cho, Y.-W. Kim, B. Kim, L.-G. Lee and B. Park, Angew. Chem. Int. Ed., 2003, 42, J. Cho, J. -G. Lee, B. Kim and B. Park, Chem. Mater., 2003, 15, J. Cho, H. Kim and B. Park, J. Electrochem. Soc., 2004, 151, A H. Lee, M. G. Kim and J. Cho, Electrochem. Commu., 2007, 9, K. Xu, M. S. Ding, S. Zhang, J. Allen and T. R. Jow, J. Electrochem.Soc., 2002, 149, A S. C. Narang, S. C. Ventura, B. J. Dougherty, M. Zhao, S. Smedley and G. Koolpe, U.S. Patent 5,830,600, X. Wang, E. Yasukawa and S. Kasuya, J. Electrochem. Soc, 2001, 148, A N. Takami, T. Ohsahi, H. Hasebe and M. Yamamoto, J. Electrochem.Soc., 2002, 149, A9. 18 C. Buhrmester, J. Chen, L. Moshurchak, J. Jiang, R. L. Wang and J. R. Dahn, J. Electrochem. Soc., 2005, 152, A H. Lee, J. H. Lee, S. Ahn, H.-J. Kim and J.-J. Cho, Electrochem. Solid-State Lett., 2006, 9, A H. Lee, S.-K. Chang, E.-Y. Goh, J.-Y. Jeong, J. H. Lee, H.-J. Kim, J.-J. Cho and S.-T. Hong, Chem. Mater., 2008, 20, S. W. Song, G. V. Zhuang and P. N. Ross, Jr., J. Electrochem. Soc., 2004, 151, A K. Matsumoto, R. Kuzuo, K. Takeya and A. Yamanaka, J. Power Sources, 1999, 81 82, G. V. Zhuang, G. Chen, J. Shim, X. Song, P. N. Ross and T. J. Richardson, J. Power Sources, 2004, 134, H. S. Liu, Z. R. Zhang, Z. L. Gong and Y. Yang, Electrochem. Solid- State Lett., 2004, 7, A K. Kang, C. -H. Chen, B. J. Hwang and G. Ceder, Chem. Mater., 2004, 16, M. Noh, Y. Lee and J. Cho, J. Electrochem. Soc., 2006, 153, A J. Kim, Y. Hong, K. S. Ryu, M. G. Kim and J. Cho, Electrochem. Solid-State Lett., 2006, 9, A J. Cho, Y. J. Kim, T.-J. Kim and B. Park, Chem. Mater., 2001, 12, S. E. Shadle, J. E. Penner-Hahn, H. J. Schugar, B. Hedman, K. O. Hodgson and E. I. Solomon, J. Am. Chem. Soc., 1993, 115, K. B. Musgrave, C. E. Laplaza, R. H. Holm, B. Hedman and K. O. Hodgson, J. Am. Chem. Soc., 2002, 124, X. Ottenwaelder, A. Aukauloo, Y. Journaux, R. Carrasco, J. Cano, B. Cervera, I. Castro, S. Curreli, M. Carmen Munoz, A. L. Rosello, B. Soto and R. Ruiz-Garcia, Dalton Trans., 2005, Y.-G. Lee and J. Cho, Electrochim. Acta, 2007, 52, This journal is ª The Royal Society of Chemistry 2008 J. Mater. Chem., 2008, 18,