Effect of Residual Lithium Compounds on Layer Ni-Rich Li[Ni_0.7Mn_0.3]O₂

First, (Ni_0.7Mn_0.3)(OH)₂ precursor was prepared via co-precipitation following our prior works.¹³ An aqueous solution of Ni(II)SO₄·6H₂O (Samchun) and Mn(II)SO₄·5H₂O (Junsei) (molar ratio of Ni:Mn = 7:3) with concentration of 2.0 mol L⁻¹ was dropped into a continuously stirred tank reactor (CSTR) with 1000 rpm under N₂ atmosphere. At the same time, NaOH solution (2.0 mol L⁻¹) and NH₄OH solution (3.6 mol L⁻¹) as chelating agents were also separately added into the reactor by adjusting the solution pH to 11.7 at 50°C. Then, the precipitated particles were filtered, washed, and dried in an oven at 80°C for 24 h. The obtained (Ni_0.7Mn_0.3)(OH)₂ was heated to produce dehydrated (Ni_0.7Mn_0.3)₃O₄ at 500°C for 5 h. The dehydrates were thoroughly mixed with an appropriate amount of LiOH and calcined at 900°C for 15 h in air.

To prepare the environment in which the as-synthesized active materials present with residual lithium compounds, the surfaces of active materials were intentionally modified by excess amount of lithium isopropoxide; for 10 g of active material, the added amount of lithium isopropoxide is 0.7 g, which corresponds to ∼0.01 mol of lithium isopropoxide. According to our preliminary experiment, addition of more lithium isopropoxide resulted in low capacity due to the thick coating layer and less amount of lithium isopropoxide below 0.007 mol did not show the presence of uniform Li₂O coating layer. Lithium isopropoxide was dissolved in anhydrous ethanol, and the as-synthesized Li[Ni_0.7Mn_0.3]O₂ was poured into the solution and stirred until the solution completed evaporation of the solvent at 80°C in air. Finally, the dried powder was heated at 400°C for 5 h in air to form lithium oxide (approximately 0.0045 mol per 1 mol Li[Ni_0.7Mn_0.3]O₂ active material) on the surface of active materials.

X-ray diffractometry (XRD; Rint-2000, Rigaku) and transmission electron microscopy (HR-TEM; JEM-3010, JEOL) were employed to characterize the synthesized powders. Time-of-flight secondary ion mass spectroscopy (ToF-SIMS; PHI TRIFT V, ULVAC-PHI) was also used to confirm the presence of the lithium oxide layer on the surface of Li[Ni_0.7Mn_0.3]O₂ powders. Also, X-ray photoelectron spectroscopy (XPS; PHI 5600, Perkin-Elmer) measurements were conducted on the surfaces of bare and coated Li[Ni_0.7Mn_0.3]O₂ materials. Macro-mode (3 mm × 3 mm) Ar-ion etching was done to analyze the chemical state and surface composition of the coated powders. The etching rate was estimated as 4 Å min⁻¹ for silica.

For the fabrication of cathodes, the as-synthesized and the Li₂O-coated Li[Ni_0.7Mn_0.3]O₂ powders were mixed with conductive materials (Super-P and Ketjen black, 1:1 in weight) and polyvinylidene fluoride in a weight ratio of 8:1:1 in N-methyl-2-pyrrolidone. The obtained slurries were applied onto Al foil and dried at 80°C in air. The electrodes were transferred to a vacuum oven and dried at 110°C overnight prior to use. Coin-type cells (2032) were assembled using Li metal (Honjo chemical) as the anode in an argon-filled glove box. The electrolyte used was 1M LiPF₆ in ethyl carbonate–dimethyl carbonate (3:7 in volume, PanaX). The cells were charged and discharged between 3.0 and 4.3 V by applying a constant current (50 mA g⁻¹) for electrochemical tests at room temperature.

In situ XRD data were obtained using an Empyrean diffractometer (PANalytical) equipped with monochromated Cu Kα radiation from 2θ = 15 to 70°, with a step size of 0.03° and a count time of 7s using a lab-made in situ cell.¹⁴ The cell was charged and discharged with a current density of 50 mA g⁻¹. The lattice parameters were calculated by the following method: the positions of the individual peaks were fitted with a psuedo-Voigt or Lorentz function, and typically 8 or 9 peak positions were input to fit the program which minimizes the least squares difference between the calculated and measured peak positions by adjusting the lattice constant and the vertical displacement of the sample.

To identify the presence of byproducts on the surface of the active materials after cycling, the cycled active materials were measured using a ToF-SIMS surface analyzer operated at 10⁻⁹ Torr, equipped with a liquid Bi⁺ ion source and pulse electron flooding. During the analysis, the targets were bombarded by the 10 keV Bi⁺ beams with a pulsed primary ion current varying from 0.3 to 0.5 pA. The total collection time was 100s and rastered over a 120 μm × 120 μm dimension.

Figure 1 shows Rietveld refinement results of bare and lithium oxide-coated Li[Ni_0.7Mn_0.3]O₂ (approximately 0.0045 mole per 1 mole Li[Ni_0.7Mn_0.3]O₂ active material). The XRD patterns could be identified as an α-NaFeO₂ structure with space group. Though the Li[Ni_0.7Mn_0.3]O₂ was modified by foreign layers, there are no significant differences in the XRD pattern compared to that of the bare Li[Ni_0.7Mn_0.3]O₂. The calculated lattice parameters also indicate no change in the crystal structure before and after the coating (Tables I and II). The amount of Ni²⁺ occupying Li site (3b), due to the similarity of ionic radius for Li⁺ (0.76 Å, coordination number: 6)¹⁵ and Ni²⁺ (0.69 Å, coordination number: 6)¹⁵ appears to be identical for both materials, approximately 5.9% in Li site, resulting in [Li_0.941Ni_0.059][Ni_0.7Mn_0.3]O₂. This is reasonable that the coating layer is comprised of low crystalline product since the heating temperature of 400°C is too low to form crystalline phase but sufficient to remove isopropoxide ingredient, resulting in formation of low crystalline Li₂O coating layer. And it is also thought that the lithium oxide does incorporate into neither Li nor Ni site because of no change in the lattice parameters (Tables I and II) but resides only on the surface of the active materials. For convenience, in the following discussion, the phases are still associated with their nominal chemical formula Li[Ni_0.7Mn_0.3]O₂, whatever the real cationic distribution.

Table I. Rietveld refinement results of XRD data for bare Li[Ni_0.7Mn_0.3]O₂.

Nominal Formula				Li[Ni0.7Mn0.3]O2
Crystal system				Rhombohedral
Space group
Atom	Site	x	y	z	g	B #x002F; Å2
Li	3a	0	0	1/2	0.941(2)	1.1
Ni2	3a	0	0	1/2	0.059(2)	1.1
Ni1	3b	0	0	0	0.7	1.0
Mn	3b	0	0	0	0.3	1.0
O	6c	0	0	0.258(3)	1	0.8
a-axis / Å				2.8804(2)
c-axis / Å				14.2543(10)
Rwp /%				14.2
Rp /%				9.15

Table II. Rietveld refinement results of XRD data for Li₂O-coated Li[Ni_0.7Mn_0.3]O₂.

Nominal Formula				Li[Ni0.7Mn0.3]O2
Crystal system				Rhombohedral
Space group
Atom	Site	x	y	z	g	B / Å2
Li1	3a	0	0	1/2	0.941	1.1
Ni2	3a	0	0	1/2	0.059(2)	1.1
Ni1	3b	0	0	0	0.7	1.0
Mn	3b	0	0	0	0.3	1.0
O	6c	0	0	0.258(2)	1	0.8
a-axis / Å				2.8808(2)
c-axis / Å				14.2548(9)
Rwp /%				11.8
Rp /%				7.46

Zoom In Zoom Out Reset image size

To confirm morphology of the as-synthesized and the Li₂O-coated coated materials, the powders were observed by SEM and TEM (Figure 2). In the SEM images (Figs. 2a and 2b), the smooth surface of active materials (Fig. 2a) is covered by flake-like particles after the Li₂O coating (Fig. 2b). From TEM observation (Fig. 2c), even the bare materials display some sediment of foreign materials on the outer surface, presumably lithium compounds such as Li₂O, Li₂CO₃ and LiOH derived from the excessive amount of LiOH addition for compensation of lithium evaporation to minimize the cation mixing during calcination. Such foreign layers are readily noticed on the surface of the active material after the Li₂O coating (Fig. 2d); the newly formed thicker layers range from 10−40 nm in thickness for the Li[Ni_0.7Mn_0.3]O₂, and the resulting coating layers appear inhomogeneous but porous showing light contrast compared to the active material.

Zoom In Zoom Out Reset image size

Residual lithium oxide, Li₂O, on the surface of active materials is readily combined with oxygen, carbon dioxide, and water molecules, and forms compounds like LiOH and Li₂CO₃, as confirmed in ToF-SIMS spectra (Figs. 3a and 3b). In particular, the Li₂O-coated sample represents stronger spectrum intensity for the Li₂C⁺ (m = 31.01) and Li₂OH⁺ (m = 31.03) fragments (Fig. 3b). The results indicate that the outer surface of Li₂O coating layer was obviously transformed to LiOH and Li₂CO₃ layers after exposure in air. Hence, the bare and Li₂O-coated materials were analyzed by pH, LiOH, and Li₂CO₃ titrations (Fig. 3c). In the result of pH titration, the coated material by Li₂O exhibits higher pH value than that of the bare. And through the LiOH and Li₂CO₃ titrations, it is confirmed that the Li₂O coating induces more formation of LiOH and, in particular, Li₂CO₃ on the surface of active materials (Scheme 1).

Zoom In Zoom Out Reset image size

Scheme 1. Surface change of Li[Ni_0.7Mn_0.3]O₂ materials after exposure in air.

The bare and the Li₂O-coated Li[Ni_0.7Mn_0.3]O₂/Li cells were continuously cycled at a constant current density of 50 mA g⁻¹ at 25°C. The first discharge capacity of the bare material appears evidently higher than that of Li₂O-coated electrode (Fig. 4a). In addition, the coulombic efficiency of the bare electrode at the first cycle is approximately 75.3%, whereas the coated material shows lower columbic efficiency (67.7%) than that of the bare. The presence of thick and non-uniform coating layers would cause lower discharge capacity and efficiency for the Li₂O-coated Li[Ni_0.7Mn_0.3]O₂ electrode. Decomposition or oxidation of those excessive lithium compounds derived from the Li₂O coating is also considered for the low coulombic efficiency at the first cycle. The capacity retentions are approximately 76% for the bare and 70% for the Li₂O-coated electrodes (Figs. 4b-4d).

Zoom In Zoom Out Reset image size

As observed in Fig. 4, deterioration of the electrode performance is obvious in the virtual condition through the surface modification by residual lithium compounds. Since the only difference between two electrodes is the presence of residual lithium compounds on the surface of active materials, the remained lithium compounds may affect structural variation during Li⁺ insertion and extraction. Hence, the bare and Li₂O-coated materials were examined by in situ XRD measurement to follow the structural changes during the first cycle (Fig. 5). In fact, both electrodes undergo similar structural evolution during the charge and discharge. Some peaks gradually shift toward higher angle in 2θ on charge that is due to structural changes from H1 (hexagonal) to H2 (hexagonal phase) phase via the intermediate M (monoclinic) phase.^16–18 And the peaks return to the original peak position on discharge without appearance of other phases, occurring in a simple topotactic reaction (Figs. 5a and 5b).

Zoom In Zoom Out Reset image size

Interlayer distances were calculated from the in situ XRD data shown in Fig. 5, and the results are plotted in Fig. 6. The resulting distances increase gradually during the charge because of the increase in coulombic (electrostatic) repulsion force between the oxygen-oxygen layers by ionicity of metal-oxygen bonding.¹⁹ Although the delivered capacity was small for the Li₂O-coated electrode, the interlayer distances vary similarly to those of the bare electrode. From this point of view, it is possible that the residual lithium compounds do not affect structural change but do interrupt Li⁺ diffusion because of the thickness and insulating properties the lithium compounds. For the reasons, the coated electrode would deliver less capacity than the bare material.

Zoom In Zoom Out Reset image size

The extensively cycled cells were disassembled, and both the bare and Li₂O-coated electrodes were washed with salt-free dimethyl carbonate solution and subsequently dried in vacuum oven at 80°C. In the TEM image of bare material, the sediments were, in part, detected on the surface of active material after the cycling, and the original smooth surface line of the particle was also kept (Fig. 7a). The coating layer on the surface of the coated material became thicker than before cycling (Fig. 7b). It is likely that the coated lithium compounds layers are altered through reaction with electrolyte during the cycling.

Zoom In Zoom Out Reset image size

For the above reason, the extensively cycled bare and coated electrodes were analyzed by ToF-SIMS to unveil how the outermost original lithium compounds coating layers were altered. Four fragments are found in the mass range of 30.6–31.4: P⁺ (m = 30.97): CF⁺ (m = 30.99), Li₂C⁺ (m = 31.01), and Li₂OH⁺ (m = 31.03). P⁺ is derived from remained LiPF₆ salt and CF⁺ is due to the presence of PVDF binder in the electrode. The development of Li₂C⁺ intensity is evidently noticed after the cycling (Figs. 8a and 8b), compared to those fresh electrodes (Figs. 3a and 3b). In comparison with the intensity of the Li₂OH⁺ fragment, the intensity of Li₂C⁺ appears to be close to that of Li₂OH⁺ for the bare electrode and the signal is somehow stronger than that of Li₂OH⁺ for the Li₂O-coated electrode. This result indicates that Li₂CO₃ layer resides just above the LiOH layer, as follows:

Zoom In Zoom Out Reset image size

Aurbach's group²⁰ suggested that CO and CO₂ are anodic oxidation products of ethylene carbonate and dimethylene carbonate, which are used as solvents in the present study. Through the above reaction, these oxidation byproducts are deposited on the outer surface of residual LiOH layer forming the Li₂CO₃ layer but leaving water molecules (eq. 1). As is clear from the XPS results, more amount of Li₂CO₃ is formed on the outermost surface of the Li₂O-coated electrode after the extensive cycling test (Figs. 8e and 8f). Hence, the relative intensity of Li₂OH⁺ diminishes compared to the fresh state. Etching the electrode surface by Bi⁺ for 50s, the carbonate layer almost disappears, and only the LiOH layer is observed for both electrodes, meaning that the formed Li₂CO₃ layer is not thick.

Commonly, electrolyte using LiPF₆ salt, which is unstable in the presence of H₂O molecules and at high temperature or high operation voltage, is decomposed by the following equations:^21,22

and

The produced LiF is generally deposited on the surface of active material. The LiF compound is not found on the XPS spectra for both fresh electrodes (Figs. 9a and 9c). After both electrodes were extensively cycled, the LiF is detected at 685 eV (Figs. 9b and 9d). In particular, the relative intensity of LiF binding energy is significantly higher for the coated material than that of bare one. It is believed that the formation of LiF is highly dependent on the amount of presenting water molecules in the electrolyte as expected from the above reactions. In this experiment, the difference is the presence of Li₂O coating layer on the surface of active material. Through the anodic decomposition of electrolytic solvent, the produced CO and CO₂ tend to bond with the hydrated Li₂O layer, namely LiOH, forming Li₂CO₃ layer in the outermost surface, and this subsequently accompanies the formation of water molecules. In comparison with the bare material, the formed Li₂CO₃ layer seems thicker for the coated electrode, which also means generation of more amounts of water molecules in the electrolyte. This accelerates the decomposition of electrolytic salt, LiPF₆, to lead to more formation of insulating LiF in the outermost surface. Therefore, the XPS result indicates the further development of LiF intensity for the Li₂O-coated electrode after the cycling test. These layers, such as LiF, Li₂CO₃, and LiOH, coexist on the surface of the active materials, so that they impede the diffusion of lithium ions due to their insulating properties and thus deterioriate the electrochemical performances of active materials as described in Scheme 2. These series of sequential occurrence are affected by the presence of residual Li₂O. Therefore, efforts should be made to minimize the amount of impurities on the surface of active materials to reduce these side reactions occurring in the interface of active materials and electrolyte during cycling. Surface modification capturing the residual Li₂O compound would be a possible way to reduce the level of the residual lithium compounds on the surface of Ni-rich layer compounds.

Zoom In Zoom Out Reset image size

Scheme 2. Effect of the residual lithium on the surface of Li[Ni_0.7Mn_0.3]O₂.

The effect of residual lithium on the surface of Li[Ni_0.7Mn_0.3]O₂ is investigated in an artificial environment provided by Li₂O coating. Exposure of the coated materials in air induces notable increase in the amount of LiOH as a major component and Li₂CO₃ as a minor component. Because of these contaminations, the surface-modified electrode shows unfavored electrode performances such as low capacity, large irreversible capacity, and poor capacity retention. It is found that the LiOH layer on the Li₂O is transformed to Li₂CO₃ layers accompanied by the formation of water. This water molecule also accelerates the decomposition of LiPF₆ salt, and this consequently gives rise to the formation of LiF toward the outermost surface of the active material. The progressive formation of insulating LiF layer impedes lithium diffusion and deteriorates the electrode performance during cycling. Therefore, we suggest that the concentration of surface lithium should be minimized to avoid these unfavorable effects.

The authors thank Miwa Watanabe, Iwate University, for her helpful assistance in the experimental work. This work was partially supported by the IT R&D program of MKE/KEIT [10041856] and the secondary battery R&D program for leading green industry of MKE/KEIT [10041094].