Abstract

The electrochemical performance of Li(Mn, M)PO4 (M = Co2+/3+, Ni2+/3+) was investigated with regard to the particle morphology. Within a controlled chemical composition, Li(Mn0.92Co0.04Ni0.04)PO4, the resultant cathode exhibited somewhat spherical-shaped nanocrystalline particles and enhanced Li+-ion storage, which was even better than the undoped LiMnPO4, up to 16% in discharge capacity at 0.05 C. The outstanding electrochemical performance is attributed to the well-dispersed spherical-shaped particle morphology, which allows the fast Li+-ion migration during the electrochemical lithiation/delithiation process, especially at high current density.

1. Introduction

Nowadays, demand for medium and large scale batteries has grown owing to the development of such applications as electric vehicles (EV), hybrid electric vehicles (HEV), and energy storage system (ESS) [1, 2]. Consequently, it is of great importance to develop new cathodes with greater energy and power densities for the next generation of Li ion batteries.

Although there are diverse cathode materials that meet the requirements of this demand, LiMPO4 (M = Fe, Mn, Co, and Ni) olivine compounds are being seriously considered as cathode materials for Li ion batteries because of their great structural stability and electrochemical performance. Olivine compounds do not suffer from structural rearrangement by the lithiation and delithiation of Li+-ions, based on the tetrahedral polyanions; the strong covalent bonds between the oxygen and the P5+ ions stabilize the three-dimensional framework of the olivine structure [3]. Therefore, LiMPO4 electrodes have substantially better stability and capacity retention along with continued cycling compared with other transition-metal oxide cathode materials such as LiCoO2, LiNiO2, LiMnO2, and LiMn2O4 [4, 5]. Moreover, due to the excellent cycle life, high capacity, and thermal stability of olivine-type materials, these materials have been substantially investigated as promising electrode materials for rechargeable Li ion batteries [68]. Among the diverse olivine-type materials, LiFePO4 has already been used as commercial cathode material in Li ion batteries. However, LiFePO4 suffers from a critical problem: it has a relatively low operating voltage region (3.4 V versus Li/Li+), which results in limited energy density. Consequently, LiMnPO4 as a representative of LiMPO4 has attracted much attention for its high operating voltage region (4.1 V versus Li/Li+), accompanied with high-energy density (697 Wh Kg−1) [9, 10]. Despite this advantage, the inherent low electronic conductivity (<10−10 S cm−1) is one of the main causes that degrades the electrochemical performance of LiMnPO4, which is manifested in poor cycle life and poor rate capability [11]. To overcome this limitation, many strategies have considered enhancing the electrochemical performance of LiMnPO4: morphology control, particle size reduction, carbonaceous material coating, and cation substitution or doping [1215].

It is generally known that downsizing electrode material to the nanoscale (50–100 nm) can remarkably improve Li+-ion accessibility and enhance the electron transport. Moreover, nanosized electrode materials effectively buffer the large volume expansion/contraction and alleviate the strain caused during repeated Li+-ion lithiation/delithiation, thereby improving the cycle life. In particular, cation doping has been proposed as one of the most effective and straightforward methods for improving the electrochemical properties of olivine-type electrode materials by improving the charge transfer ability [16]. Lee et al. also reported the enhancement of electrochemical performance of LiMnPO4 by Mg doping without increasing the rate capability [17]. In addition, Kim et al. revealed that there are differences in Fe-Co codoped LiMnPO4 and only Fe-doped LiMnPO4 composition [18].

Here, we report the enhanced electrochemical performance of LiMnPO4 according to the dopant ions and particle morphologies. Various cations were adopted in LiMnPO4 via a microwave-assisted hydrothermal process, which has advantages such as simple and fast reaction process [19, 20]. The resultant LiMnPO4 and cation-doped LiMnPO4 exhibited nanorod and spherical morphologies, respectively, and we investigated their electrochemical performances along with the origins of their different properties. Different cations, such as Ni and Co, were applied into the LiMnPO4 lattice. In particular, we prepared LiMn0.9Co0.1PO4, LiMn0.9Ni0.1PO4, and LiMn0.92Co0.04Ni0.04PO4 and investigated their performance as electrode materials in relation to charge-discharge curves, cycling stability, and rate capability.

2. Experiment

Li()PO4 (M = Ni, Co) was prepared by a microwave-assisted hydrothermal method. The raw materials were Li(CH3COO)·2H2O (98%, Sigma Aldrich), MnSO4·H2O (99%, Sigma Aldrich), CoSO4·7H2O (99%, Sigma Aldrich), NiSO4·7H2O (99%, Sigma Aldrich), CTAB (cetyltrimethylammonium bromide), and H3PO4 (85%, High Purity Chemicals). The CTAB was used as a surfactant. In the synthesis of Li()PO4 (M = Ni, Co), the molar ratio of Li  : (Mn, M) : P was 3 : 1 : 1, and then it was dissolved in 50 mL of 2-methoxy ethanol with vigorous stirring. Then, the cationic surfactant CTAB was dissolved in the above solution. The homogeneously dispersed solution was transferred to a 100 mL Teflon-liner and placed in a microwave irradiation system (Model MARS6, CEM Corp., USA). The solution was heated at 260°C and maintained for 1 hour. The as-obtained products were collected, washed several times with DI water, and dried at 90°C for 24 h. For the carbon-coating on the surface of the electrode particles, LiMnPO4 powder was milled with 10 wt% of sucrose in ethanol by a planetary high-energy ball-milling (400 rpm, 3 h). After the ball-milling, the samples were put into an alumina boat in a furnace and annealed at 700°C for 1 h under reducing atmosphere (95% N2/5% H2).

The crystal structure of the products was investigated by powder X-ray diffraction (XRD) on a Rigaku Ultima IV Diffractometer using Cu K < radiation operating at 40 kV and 40 mA. The surface morphology of the synthesized particles was studied using a scanning electron microscope (FE-SEM, Tescan Mira 3 LMU FEG, 20 kV). X-ray photoelectron spectroscopy (XPS, K-Alpha Thermo Scientific) was used to characterize the chemical state of the elements.

The electrochemical performance was investigated using CR2032 coin type cells. The cathodes were prepared by blending the active materials LiMnPO4, Super-P, and polyvinylidene fluoride (PVdF, Kureha KF-1100) as a binder in a weight ratio of 70 : 20 : 10 and then dissolved in NMP (N-methyl-2-pyrrolidone, Sigma Aldrich, 99.5%) solution. The mass loading of the active material was 1.2 mg/cm2. The resulting slurry was spread on aluminum foil as a current collector (Doctor Blade method), pressed, and then vacuum-dried at 120°C for 12 h. Lithium metal was used as a negative electrode, and 1 M LiPF6 dissolved in a mixture of ethylene carbonate diethyl carbonate (EC/DEC 1 : 1 v/v) was used as electrolyte. The assembly of the coin cells was carried out in a dry argon-filled glove box.

The galvanostatic charge-discharge capacity, rate capability, and cyclic performance were tested at current densities of C/20 (1 C = 170 mA g−1) between 2.7 and 4.5 V versus Li/Li+ using a multichannel battery testing equipment (TOSCAT-3100, Toyo Co.). Cyclic voltammetry (CV) was operated on a multi-electrochemical analyzer system (Bio-Logic VSP 0417) between 3.0 and 5.0 V at a scanning rate of 0.2 mV s−1.

3. Results and Discussion

X-ray diffraction (XRD) patterns of LiMnPO4, LiMn0.9Ni0.1PO4, LiMn0.9Co0.1PO4, and LiMn0.92Ni0.04Co0.04PO4 are shown in Figure 1(a). All the samples were single-phased with the ordered olivine structure of a Pnma orthorhombic system, which is in good agreement with the standard value (JCPDS 01-072-1237). After elemental substitution, there was no deviation in the peak position or the FWHM (Full With at Half Maximum), indicating that the elemental substitution results in no alteration of the crystalline structure.

To clarify the variations in the chemical state of Ni and Co elements, X-ray photoelectron spectroscopy (XPS) measurements were carried out, and the typical XPS spectra for Co 2p and Ni 2p for the LiMn0.92Ni0.04Co0.04PO4 are shown in Figure 1(b). As shown, the binding energies were observed at 780, 796, 854, and 872 eV, which are attributed to Co 2p3/2, Co 2p1/2, Ni 2p3/2, and Ni 2p1/2, assigned well with those for Co3+/Co2+ and Ni3+/Ni2+, respectively [21, 22]. Regarding the chemical state of Ni and Co, we concluded that Co and Ni were successively doped on LiMnPO4, which may provide notable electrochemical activities of the LiMnPO4 as electrode materials.

The scanning electron micrographs (SEM) of as-obtained LiMnPO4, LiMn0.9Ni0.1PO4, LiMn0.9Co0.1PO4, and LiMn0.92Ni0.04Co0.04PO4 samples are presented in Figure 2. As can be seen, the pristine LiMnPO4 sample has a nanorod morphology with a diameter of about 40 nm, whereas all the Co-/Ni-doped LiMnPO4 samples have spherical morphologies with a diameter of about 30 nm. Since the particle properties such as morphology and size have a great effect on the performance of Li ion batteries, LiMnPO4 and Co-/Ni-doped LiMnPO4 may exhibit different electrochemical features. When the particles were scaled down to nanoscale, the surface-to-volume ratios increased, leading to the high accessibility of Li+-ions to the electrode materials during the lithiation/delithiation process. To obtain LiMnPO4 with fine nanosized particles with excellent electrochemical properties, it is essential to use a hydrothermal method and add a surfactant because the electrochemical performance of the electrode is strongly related to the particle size, which determines the diffusion length and surface area for the reaction. By adopting CTAB as a surfactant, the grain size increase can be prevented, which demonstrates that CTAB can act as an internal reducing agent for homogenizing particle size [23, 24]. Not only does the intrinsically low electronic conductivity of LiMnPO4 diminish the cathode performance, but also the anisotropic diffusion channels for Li+-ions in the olivine structure result in low ionic conductivity. For this reason, controlling particle size and morphology is essential in preparing high performance LiMnPO4 [15].

The cyclic voltammetry (CV) curves of LiMnPO4 results presented in Figure 3 confirm the electrochemical reaction at the given voltage region. The reduction and oxidation peak of LiMnPO4 are located at 3.85 and 4.35 V, respectively, indicating that the peak potential difference () is about 0.5 V at a scan rate of 0.2 mV s−1. These results confirm that the suitable oxidation/reduction reactions of the Mn2+/Mn3+ redox couple exist in the olivine structure [3, 18, 25].

The voltage profiles of the LiMnPO4 and Co-/Ni-doped LiMnPO4 at different current densities are given in Figure 4. The initial charge-discharge curves of all the samples were tested at C/20 (8.5 mA g−1) and are shown in Figure 4(a). The specific capacity of LiMnPO4 was 115 mA h g−1, which is 68% of the theoretical capacity. The problem of the olivine structure of LiMnPO4 is its low electronic and ionic conductivity, so electrochemical properties are highly dependent on the kinetics of LiMnPO4. The electrochemical properties of pure LiMnPO4 are poor due to the blockage of the diffusion path of the Li+-ions by excess Mn2+ ions on Li sites, which deteriorates the capacity [26]. Furthermore, the electron transport is restricted by the strong Mn–O bonds, and the Li+-ions diffusion is limited by the Li–O bonds. Thus, to improve the electrochemical properties of the samples obtained using the elemental substitution, we selected the transition metals Ni and Co. The initial discharge capacity of LiMnPO4, LiMn0.9Ni0.1PO4, LiMn0.9Co0.1PO4, and LiMn0.92Ni0.04Co0.04PO4 was 115, 136, 121, and 134 mA h g−1, respectively. At 10 C (1700 mA g−1), the discharge capacity was 57 mA h g−1 for LiMnPO4 (49% of the capacity at C/20), whereas the discharge capacity of LiMn0.92Ni0.04Co0.04PO4 was 88 mA h g−1 (66% at C/20). Clearly, the Co-/Ni-doped LiMnPO4 achieved a higher capacity than pristine LiMnPO4. The reason for the increment in the electrochemical properties of LiMnPO4 is that the substitution of a small amount of Co and Ni in LiMnPO4 led to a reduction in volume changes during the delithiation of Li+-ions. Furthermore, the small binding energy of Li–O bonds makes it easy to migrate Li+-ions [6, 27, 28]. Moreover, the pristine LiMnPO4 exhibited a nanorod morphology, which has one-dimensional ionic diffusion paths, and the Co-/Ni-doped LiMnPO4 showed a deviation in diffusion paths due to its spherical morphologies. Chen et al. experimentally demonstrated the movement of Li+-ions in the olivine structure by studying the phase boundary through TEM [29]. The Li+-ions migrated along the direction, and it was also reported by Morgan et al. (or coworkers) that the LiMPO4 with a nanorod morphology had a one-dimensional diffusion path along the direction based on the space group of Pnma [30, 31]. As mentioned above, since one of the major constraints on the conductivity of Li+-ions is that the diffusion channel of Li+-ions is one-dimensional, we can infer that the diffusion of Li+-ions may be alleviated by extending the diffusion paths. In particular, the lattice enlargement along the direction facilitates the expansion of Li+ diffusion channels, which sufficiently increases the mobility of Li+-ions. Therefore, we assumed that the lattice parameter in the direction was somewhat increased by the changing particle morphology from nanorod to spherical, which resulted in mitigating the diffusion of Li+-ions in the LiMnPO4 lattice. Thus, it is noteworthy that the diffusion of Li+-ions can be improved not only by reducing the particle size to nanoscale but also by controlling the particle morphology.

To address the question of whether Co and Ni doping result in the improved electrochemical performance of LiMnPO4, we measured and compared the rate capability of all electrodes and present the results in Figure 5. The measurement was carried out with an increasing discharge current density at C/20~10 C, and then we decreased it back to C/20. It can be seen that LiMn0.92Ni0.04Co0.04PO4 exhibited a capacity of 89 mA h g−1, which is a much better rate performance than the others at a high current density of 10 C. As for the pure LiMnPO4 samples, they showed poorer behavior, 57 mA h g−1 at 10 C. These results might be understood by the proposed mechanism of Mn-sited substituted LiMnPO4 demonstrated by Wang et al. They demonstrated that the Mn-sited substituted LiMnPO4 denoted improved electronic conductivity by changing the electronic structure of LiMnPO4 in virtue of the empty orbits in dopant ions [32]. Similar to this, the electronic conductivity of Co-/Ni-doped LiMnPO4 can be enhanced by changing the electronic structure of LiMnPO4, because there are empty orbits in Co2+ (3d7) and Ni2+ (3d8), which provide impurity energy levels in the Co-/Ni-doped LiMnPO4. In this way, the kinetics of Co-/Ni-doped LiMnPO4 were substantially developed. Therefore, we concluded that the origins of enhanced kinetics for the Co-/Ni-doped LiMnPO4 were both the proper particle morphology and the changing electronic structure.

To support these results, the cycling stability of discussed samples under current density at C/10 is shown in Figure 6. The capacity retention of LiMn0.9Ni0.1PO4, LiMn0.9Co0.1PO4, and LiMn0.92Ni0.04Co0.04PO4 after 100 cycles was 90, 92, and 92%, respectively, whereas that of LiMnPO4 was 82%, which is significantly lower than the other samples. The problem of rapid capacity fading is the constantly emerged inactive material caused by the structure destruction during Li+-ions lithiation and delithiation [33, 34].

In addition, the cycling stability diminishes due to the Jahn-Teller distortion and low ionic conductivity of LiMnPO4. Many researchers have reported that the cation-doped LiMnPO4 exhibits an improvement in cycle life and rate capability. Ni and Gao reported that copper doping on LiMnPO4 resulted in improvement in the discharge capacity and cycle life [35].

4. Conclusions

We developed and synthesized olivine-type LiMnPO4 and cation-doped Li(Mn, M)PO4 (M = Co2+/3+, Ni2+/3+) electrode materials by a microwave-assisted hydrothermal method and investigated their electrochemical performance. The cation doping on the LiMnPO4 resulted in significant enhancement in the electrode performance with excellent cycle life, higher capacity, and greater rate capability compared with that of pristine LiMnPO4 due to the improved ionic and electronic conductivity. In particular, the LiMn0.92Co0.04Ni0.04PO4 acquired a substantial capacity retention of 92% after 100 cycles, exhibiting a capacity of 134 mA h g−1 at 0.05 C and 89 mA h g−1 at 10 C, whereas LiMnPO4 had a capacity retention of 82% after 100 cycles with a capacity of 115 mA h g−1 at 0.05 C and 57 mA h g−1 at 10 C. Our results clearly demonstrated the improvement in performance of LiMnPO4 through the cation doping.

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.

Acknowledgments

This work was supported by the key project of the Korea Research Institute of chemical technology. This research was also supported by Global Frontier Program through the Global Frontier Hybrid Interface Materials (GFHIM) of the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (2013M3A6B1078874).