Structural and electrochemical investigation of LiNi_0.8Co_{0.2 − x} M _x O₂ (M = Al, Al+Mg, Al+Mg+Fe) synthesized by solid-state method

Abstract

Lithium nickel cobalt oxide materials doped with Al, Mg, and Fe were synthesized by solid-state reaction at 800 °C for 18 h to study the effects of adding transition and nontransition metals to the structure. Crystalline compounds were obtained as revealed by powder X-ray diffraction (XRD). Energy dispersive analysis of X-rays (EDAX) was used to determine the elemental ratio of all the samples. Impedance measurements showed that all samples have decreasing conductivities at higher temperatures and gave negative activation energies. The addition of nontransition metals actually decreased the conductivities of the materials.

Introduction

Lithium nickelate [1–3] is becoming one of the most promising positive electrode materials for lithium-ion batteries. Research in cathode materials for rechargeable lithium batteries has a great deal of interest on the layered LiNiO₂ and LiCoO₂ materials in recent years [4–6]. Doped LiNiO₂ may possess better electrochemical properties than LiNiO₂ or LiCoO₂. Single-phase LiNi_0.8C_0.2O₂ material was prepared by Saadoune and Delmas at 800 °C [7]. Charge–discharge cycling of LiNi_1 − y Al _y O₂ as positive electrode material in lithium cells has shown that aluminum substitution suppresses all the phase transitions observed for the LiNiO₂ system [10].

According to Fujita et al. [8] LiNi_0.85Co_0.15O₂ can be prepared at temperatures as low as 400 °C. Recently, Madhavi et al. [9] have examined the cathodic behavior of aluminum substituted phases, LiNi_0.7Co_{0.3 − z} Al _z O₂ (0 ≤ z ≤ 0.20) and established that the composition with z = 0.05 showed the optimum behavior in which it retains 70% of the initial discharge capacity at the end of 100 charge–discharge cycles [9]. Therefore, the combined Co and Al substitution for nickel appears very promising because it allows the positive effect of cobalt with improvements of the lamellar character of the structure and the electrochemical properties [10].

A decrease in the nickel ions in the interslab space was observed by cobalt substitution, leading to an improvement of the electrochemical properties [11–15]. Previous studies done on the LiNi_{1 − y − z} Co _y Al _z O₂ materials confirm the great interest in their electrochemical properties and thermal stabilities of these materials [16–18]. In this study, the structural characterization and electrochemical properties of these four materials are presented: LiNi_0.8Co_0.2O₂, LiNi_0.8Co_0.1Al_0.1O₂, LiNi_0.8Co_0.1Al_0.05Mg_0.05O₂, and LiNi_0.8Co_0.1Al_0.05Mg_0.025Fe_0.025O₂.

Experimental

The single-phase cathode LiNi_0.8Co_{0.2 − x} M _x O₂ (M = Al, Al+Mg, Al+Mg+Fe), which are (C1) LiNi_0.8Co_0.2O₂, (C2) LiNi_0.8Co_0.1Al_0.1O₂, (C3) LiNi_0.8Co_0.1Al_0.05Mg_0.05O₂, and (C4) LiNi_0.8Co_0.1Al_0.05Mg_0.025Fe_0.025O₂, were prepared using the solid-state reaction method. A stoichiometric amount of LiNO₃ (formula weight = 68.94), Ni(NO₃)₂.6H₂O (formula weight = 290.81), Co(NO₃)₂·6H₂O (formula weight = 291.03), Al(NO₃)₃·9H₂O (formula weight = 375.14), Mg(NO₃)₂·6H₂O (formula weight = 256.41), and Fe(NO₃)₃·9H₂O (formula weight = 404), from Aldrich Chemicals (99.5% assay), were mixed in a porcelain crucible and calcined in a furnace at 800 °C for 18 h. The cathode materials were then left to cool to room temperature and were ground with a mortar and pestle until it becomes a fine powder.

To investigate the crystal structure, the powder obtained was analyzed by performing the X-ray diffraction (XRD) method using PHILIPS PW 1840 powder diffractometer instrument. The scan data were collected in the 2θ range of 10 < 2θ < 180°. The ratios of all elements of the materials were identified by energy dispersive analysis of X-rays (EDAX), which was attached to the scanning electron microscopy (SEM) PHILIPS MODEL 515 SEM instrument.

The composite electrodes were prepared by mixing the cathode material LiNi_0.8Co_{0.2 − x} M _x O₂ (M = Al, Al+Mg, Al+Mg+Fe) powder, suitable binding agent, and activated carbon in acetone. After a homogeneous blend was obtained, the mixture was cast onto a glass surface to form a film which was attached to an aluminum mesh current collector. Finally, the materials were cut into round shapes with diameters of 1 cm and used as the cathode. Mesocarbon microbead (MCMB) was used as the anode and 1 M LiPF₆ salt in equal volumes of ethylene carbonate (EC), dimethylene carbonate (DMC), and propylene carbonate (PC) was used as the electrolyte.

Impedance spectroscopy was done for all the materials at different temperatures by Hioki Instrument. Cyclic voltammetry was performed using AutoLab PGSTAT 30 Potentiostat–Galvanostat instrument using lithium metal as the counter electrode and the reference electrode in an electrochemical cell assembled in a dry box in an argon gas environment.

Result and discussion

Materials structure characterization

From the EDAX results shown in Table 1, it can be seen that the amount of atomic ratios of all elements in the cathodic material are agreeable with the amount of starting materials.

Table 1 Atomic ratios of elements in the materials and the amounts of chemicals used

Figures 1, 2, 3, and 4 shows the XRD patterns of (C1) LiNi_0.8Co_0.2O₂, (C2) LiNi_0.8Co_0.1Al_0.1O₂, (C3) LiNi_0.8Co_0.1Al_0.05Mg_0.05O₂, and (C4) LiNi_0.8Co_0.1Al_0.05Mg_0.025Fe_0.025O₂ materials.

Fig. 1

Power XRD for (C1) LiNi_0.8Co_0.2O₂

Fig. 2

Powder XRD for (C2) LiNi_0.8Co_0.1Al_0.1O₂

Fig. 3

Powder XRD for (C3) LiNi_0.8Co_0.1Al_0.05Mg_0.05O₂

Fig. 4

Powder XRD for (C4) LiNi_0.8Co_0.1Al_0.05Mg_0.025Fe_0.025O₂

The powder XRD diffractograms of all four materials gave sharp peaks due to good crystallinity. All the samples have similar reflection peaks at similar angles, indicating that all the materials are isostructural with LiNi_0.8Co_0.2O₂ and, therefore, can be indexed in the R3m space group [7]. $\left( {\frac{{I_{003} }}{{I_{104} }}} \right)$ intensity ratio plays a part in showing the characterization of an ordered or disordered lamellar phase. The $\left( {\frac{{I_{003} }}{{I_{104} }}} \right)$ ratio >1.0 is an indication of highly ordered lamellar phase with no cation mixing and vice versa [19]. All four samples gave different peak intensities. The relative intensity ratio of the $\left( {\frac{{I_{003} }}{{I_{104} }}} \right)$ peaks for (C1) LiNi_0.8Co_0.2O₂, (C2) LiNi_0.8Co_0.1Al_0.1O₂, and (C3) LiNi_0.8Co_0.1Al_0.05Mg_0.05O₂ is 1.3, 1.3, and 1.0, respectively, and show no cation mixing.

The (C4) LiNi_0.8Co_0.1Al_0.05Mg_0.025Fe_0.025O₂ has a ratio of 0.52 and shows cation mixing. When the $\left( {\frac{{I_{003} }}{{I_{104} }}} \right)$ ratio is more than 1.0 the nontransition metals like Al and Mg will occupy the Ni and Co sites only, whereas if the ratio is smaller than 1.0, the nontransition metals will occupy the lithium sites, too [20]. Chang et al. [21] have indicated that Mg²⁺ ions occupy only the Ni and Co sites in the structure of the Li(Ni_0.8Co_0.11Mg_0.09)O₂ material, and Madhavi et al. [9] have shown that Al³⁺ ions occupy only the Ni and Co sites in the Li(Ni_0.7Co_{0.3 − z} Al _z )O₂ structure. Both materials have the intensity ratio $\left( {\frac{{I_{003} }}{{I_{104} }}} \right)$ of more than 1.0. It was also stated that with cation mixing in the materials, the Mg²⁺ or Ni²⁺ ions may occupy the lithium sites in the lattice [9].

For the (C4) LiNi_0.8Co_0.1Al_0.05Mg_0.025Fe_0.025O₂ material, based on the $\left( {\frac{{I_{003} }}{{I_{104} }}} \right)$ ratio, it can be concluded that there are some Mg²⁺ or Al³⁺ ions which occupy the lithium sites. The relative peak intensities of (006) and (101), $\left( {\frac{{I_{006} }}{{I_{101} }}} \right)$ for (C1) LiNi_0.8Co_0.2O₂, (C2) LiNi_0.8Co_0.1Al_0.1O₂, (C3) LiNi_0.8Co_0.1Al_0.05Mg_0.05O₂, and (C4) LiNi_0.8Co_0.1Al_0.05Mg_0.025Fe_0.025O₂ is 0.32, 0.64, 0.87, and 0.79, respectively. The splitting of the (006) and (102) peaks and the (108) and (110) peaks had occurred in the (C1) LiNi_0.8Co_0.2O₂ and (C2) LiNi_0.8Co_0.1Al_0.1O₂ materials, but the splitting is not clearly visible in (C3) LiNi_0.8Co_0.1Al_0.05Mg_0.05O₂ and (C4) LiNi_0.8Co_0.1Al_0.05Mg_0.025Fe_0.025O₂.

Electrochemical studies

Electrical conductivity and thermoelectrical power measurement evidenced a gradual change in the electronic properties from electron localization to electron delocalization upon lithium deintercalation [22]. A strong increase in the conductivity and the transition from semiconductor behavior to metallic behavior were reported, as a consequence of the creation of holes in the t ₂ band upon lithium deintercalation [22]. According to Menestrier and colleagues [24], the metal–nonmetal transition was the driving force for the existence of the biphasic domain, in agreement with the experimental results of Imanishi et al. [23]. Pseudometallic behavior is observed, as conductivity increases with decreasing temperature [22].

Impedance spectroscopy was done on all four materials. Figure 5 shows the Nyquist plots taken at different temperatures for the (C1) LiNi_0.8Co_0.2O₂ material. It gave a conductivity of 1.18 × 10⁻³ S cm⁻¹ for the lowest temperature of 300 K and conductivity decreases with increasing temperature. Nyquist plots of the (C2) LiNi_0.8Co_0.1Al_0.1O₂, (C3) LiNi_0.8Co_0.1Al_0.05Mg_0.05O₂, and (C4) LiNi_0.8Co_0.1Al_0.05Mg_0.025Fe_0.025O₂ materials were obtained at different temperatures and are shown in Figs. 6, 7, and 8, respectively. The conductivity at 300 K for the (C2) LiNi_0.8Co_0.1Al_0.1O₂ and (C3) LiNi_0.8Co_0.1Al_0.05Mg_0.05O₂ material is 5.0 × 10⁻⁴ and 3.51 × 10⁻⁴ S cm⁻¹, respectively.

Fig. 5

Nyquist plots taken at different temperatures for the (C1) LiNi_0.8Co_0.2O₂ material

Fig. 6

Nyquist plots taken at different temperatures for the (C2) LiNi_0.8Co_0.1Al_0.1O₂ material

Fig. 7

Nyquist plots taken at different temperatures for the (C3) LiNi_0.8Co_0.1Al_0.05Mg_0.05O₂ material

Fig. 8

Nyquist plots taken at different temperatures for the (C4) LiNi_0.8Co_0.1Al_0.05Mg_0.025Fe_0.025O₂ material

The (C4) LiNi_0.8Co_0.1Al_0.05Mg_0.025Fe_0.025O₂ material exhibits the lowest conductivity of 3.03 × 10⁻⁴ S cm⁻¹ for all temperatures. Variations of the logarithm of the electrical conductivity (S cm⁻¹) vs. reciprocal temperature (K⁻¹) of all four materials are shown in Fig. 9. Figure 10 shows the comparisons in electrical conductivities at each temperature to examine the differences in each material. Table 2 gives the activation energy of the materials from the conductivity measurement calculated from the Arrhenius equation:

$$\ln k = \ln A - \frac{{E_A }}{{{\text{RT}}}}$$

Fig. 9

Log (k/S cm⁻¹) vs. the reciprocal of temperature 1,000 T⁻¹/K⁻¹

Fig. 10

The variation of conductivity k/S cm⁻¹ with temperature for each sample

Table 2 Activation energies for all four cathodic materials

From Fig. 10, the highest conductivity obtained was for (C1) LiNi_0.8Co_0.2O₂ for all temperatures and also showed larger variation of conductivity with temperature. With the introduction of nontransition metals such as Al and Mg, the conductivities of the materials decreased for each temperature and showed smaller variations with temperatures.

The conductivity and temperature variation studies suggest that the materials experience phase transformation from semiconductor to pseudometallic phase, as also been suggested by other workers [22] because metal conductivity also decrease with temperature increase.

Cyclic voltammetry was done to examine the oxidation–reduction peaks of the materials. The oxidation peak had occurred at about 4.5 V and the reduction peak at 2.9 V while a smaller oxidation peak can be seen at 2.3 V for the 0.1 mV s⁻¹ scan rate for the (C1) LiNi_0.8Co_0.2O₂ material shown in Fig. 11.

Fig. 11

Cyclic voltammogram for the (C1) LiNi_0.8Co_0.2O₂ material at 1 × 10⁻⁴ V s⁻¹

Figure 12 shows the voltammogram for the (C2) LiNi_0.8Co_0.1Al_0.1O₂ material at 1 × 10⁻⁴ V s⁻¹ scan rate and have one oxidation–reduction pair of peaks at around 2.4 and 2.6 V, respectively.

Fig. 12

Cyclic voltammogram for the (C2) LiNi_0.8Co_0.1Al_0.1O₂ material at 1 × 10⁻⁴ V s⁻¹

The voltammograms for the (C3) LiNi_0.8Co_0.1Al_0.05Mg_0.05O₂ and (C4) LiNi_0.8Co_0.1Al_0.05Mg_0.025Fe_0.025O₂ materials are shown in Figs. 13 and 14, respectively. From Fig. 13, the cyclic voltammogram at 1 × 10⁻⁴ V s⁻¹ for (C3) LiNi_0.8Co_0.1Al_0.05Mg_0.05O₂ gave oxidation–reduction peaks at 5.0 and 4.8 V, respectively. While in Fig. 14, the cyclic voltammogram at 1 × 10⁻⁴ V s⁻¹ for (C4) LiNi_0.8Co_0.1Al_0.05Mg_0.025Fe_0.025O₂ gave oxidation–reduction peaks at 6.5 and 6.0 V, respectively.

Fig. 13

Cyclic voltammogram for the (C3) LiNi_0.8Co_0.1Al_0.05Mg_0.05O₂ material at 1 × 10⁻⁴ V s⁻¹

Fig. 14

Cyclic voltammogram for the (C4) LiNi_0.8Co_0.1Al_0.05Mg_0.025Fe_0.025O₂ material at 1 × 10⁻⁴ V s⁻¹

Conclusion

The (C1) LiNi_0.8Co_0.2O₂, (C2) LiNi_0.8Co_0.1Al_0.1O₂, (C3) LiNi_0.8Co_0.1Al_0.05Mg_0.05O₂, and (C4) LiNi_0.8Co_0.1Al_0.05Mg_0.025Fe_0.025O₂ cathode materials were synthesized via solid-state method at a temperature of 800 °C for 18 h. Powder XRD of all four materials showed crystalline phase and isostructural with the (C1) LiNi_0.8Co_0.2O₂ material. Impedance spectroscopy studies gave highest conductivity for the (C1) LiNi_0.8Co_0.2O₂ material at all temperatures, and all four materials have decreasing conductivities with increasing temperatures. The addition of nontransition metals such as Al and Mg actually decreased the conductivities at all temperatures.