Long-term cycle stability at a high current for nanocrystalline LiFePO₄ coated with a conductive polymer

All chemicals and materials were obtained from the Sigma-Aldrich corporation and used without further purification. All aqueous solutions were prepared with ultra-pure water deionized by a millipore milli-Q system (18 MW cm⁻¹). LiOH (99%), FeSO₄·7H₂O (99 + %) and H₃PO₄ (85%) were used as starting materials. Citric acid, C₆H₈O₇ (99%), or a triblock copolymer (P123), poly(ethylene glycol)-block-poly(propylene glycol)-blockpoly(ethylene glycol) was used as a surfactant to control the particle growth and surface morphology.

LiFePO₄ crystalline particles were synthesized by a hydrothermal method from the stoichiometric mixture of LiOH, FeSO₄·7H₂O and H₃PO₄ (3:1:1 molar ratio). L-ascorbic acid (C₆H₈O₆, 99%) was added as a reducing agent to the stoichiometic mixture of the starting materials (Fe²⁺: C₆H₈O₆ = 4:1 molar ratio). Ethylene glycol was added into the aqueous solution to disperse the inorganic salts (1:1 volume ratio). The resulting gray solution was transferred into a 125 ml teflon-lined stainless steel autoclave and heated for 12 h at 160 °C. After the hydrothermal reaction was completed, the precipitate was completely washed and dried at 80 °C.

Conductive polymer, PEDOT, was prepared by polymerizing 3,4-ethylene dioxythiophene monomer and a dopant, para-toluene sulfonic acid (p-TSA), at a 2:1 mole ratio by using anhydrous FeCl₃ as an oxidizing catalyst [16, 17]. The PEDOT was dissolved in DMSO or cresol. PEDOT was coated on the LiFePO₄ particles by simply dispersing the LiFePO₄ powder in the conductive polymer solution at room temperature and evaporating the solvent in a vacuum oven at 60 °C.

Powder x-ray diffraction profiles were recorded on a Philips PW1050 diffractometer using graphite-monochromated CuKα radiation (λ = 1.540 56 Å). Data were collected over the 2θ range from 10° to 100° for 4 s in each 0.02° step at ambient temperature. The structural parameters were refined by using the Rietveld program FULLPROF [18]. Scanning electron microscopes (FESEM JSM-6700F and SEM JSM-6380, JEOL, Japan) were used to examine the particle morphology. The electrochemical measurements were carried out using coin cells (CR 2032 type) assembled in a dry room. The cathodes were prepared by compressing and drying after spreading a mixture slurry of the active material (LiFePO₄/PEODT coated), acetylene black (AB) and polyvinylidene fluoride (PVDF) in NMP. The electrolyte was 1 M LiPF₆ in a 1:1 mixture of ethylene carbonate (EC)/dimethyl carbonate (DMC) and the separator was a porous polyethylene film (Celgard 2300 μm). The cells were cycled with a constant current mode (galvanostatically) in a potential range of 2.0–4.3 V using a multichannel battery test system (Maccor Series 4000, USA).

The observed, calculated and difference XRD patterns from the Rietveld refinement for the as-synthesized nanocrystalline LiFePO₄ are shown in figure 1. All diffractions for the nanocrystalline LiFePO₄ in the powder XRD patterns were indexed by an orthorhombic structure with the Pnma space group with a = 10.3177 Å, b = 5.9950 Å, c = 4.6966 Å, v = 0.29051 nm³. The final atomic coordinates, isotropic temperature factors and agreement factors from this structural model are listed in table 1. In order to improve the intrinsic problem of low ion diffusivity for LiFePO₄, it is essential to minimize defect concentration and increase the crystallinity, because the Li⁺ transport is one-dimensional in olivine structure. Li⁺ mobility has been simulated to be high in the tunnels along the [010] direction only and Li⁺ hopping between tunnels is very unlikely in olivine structure [6, 7, 19–21].

Zoom In Zoom Out Reset image size

Minimizing the crystallite size down to submicron level is also essential. Various LiFePO₄ nanostructures and complex hierarchical architectures with well-defined sizes and morphologies are of great interest [5–7, 11–16]. Figure 2 shows the SEM images of the individual LiFePO₄ nanoparticles and the LiFePO₄ microstructures hierarchically assembled with the nanoplates that were synthesized by the hydrothermal method. Surfactant molecular size and its quantity used for the preparation have a significant influence on the particle morphology. The well-dispersed individual LiFePO₄ crystals formed with the P123, block copolymers, surfactant displayed rectangular- or rhombus-shapes with particle sizes that were approximately 500 nm in length, and 100–200 nm in width and in thickness (nanoplates), as shown in figure 2(a). When the surfactant was changed from P123 to citric acid, the LiFePO₄ nanoplates were hierarchically organized into microstructures. The molecular surfactant, citric acid, can be adsorbed onto particle surfaces during the particle growth, and in turn can organize the nanoscale building blocks into complex hierarchical architectures via a self-assembly process. The LiFePO₄ hierarchical architectures with well-defined dumbbell-like or donut-like microstructures were successfully synthesized by using citric acid as the surfactant. Dumbbell-like microstructures of LiFePO₄ with lengths of about 8 mm were hierarchically constructed with the nanoparticles (90–120 nm thick), as shown in figure 2(b). When quite an excess amount of the molecular surfactant was used, growth of the nanoplates of crystalline LiFePO₄ seemed to be inhibited. The particles were considerably thinner than those without the excess amount of the surfactant added in the reactor. The thinner LiFePO₄ nanoplates (about 50 nm thick) were hierarchically organized into donut-like microstructures, as shown in figure 2(c). Hierarchical microstructures formed through the face-to-face packing of nanoplates. The large surface of each nanoplate is considered to be (010) plane, which has the lowest surface energy calculated by Fisher and Islam [19] and Ceder et al [20, 21]. The (010) surface is normal to the most facile pathway for lithium ion conduction along the [010] channel.

Zoom In Zoom Out Reset image size

Nanocrystals have the benefit of enlarging the specific surface area with the decreased diffusion length for Li⁺ ions. In this work, these particles were coated with a conductive polymer, PEDOT, in order to increase the charge transport kinetics. Figure 3 shows a typical cyclic voltammogram of a Li cell with the cathode composed of nanocrystalline LiFePO₄ coated with PEDOT (10 wt%). The pair of the Fe²⁺/Fe³⁺ redox couple with an oxidation peak at 3.55 V and a reduction peak at 3.33 V corresponds to the charge/discharge reaction of the battery. The peak shape is almost symmetric and the potential separation between the two peaks is small. The first charge/discharge curves of the lithium batteries assembled with the LiFePO₄–PEDOT (10 wt%) coated cathodes of three different morphologies are shown in figure 4(A). All the curves show reversible lithium extraction/insertion during the charging/discharging processes with varying specific capacities. The charge and discharge profile shown in figure 4(A) demonstrates that the individual LiFePO₄ nanoparticle cathode shows the highest specific capacity of 170 mAh g⁻¹. The dumbbell- and donut-shaped microstructure LiFePO₄ cathodes show capacity values of about 120 mAh g⁻¹. The increase in the specific capacity certainly reflects the increase in the specific surface area. In addition, the smaller the particle is, the more densely the particles can be packed with smaller interstitial volumes in the cathode. The tap density of the positive electrode composed of the nanocrystalline LiFePO₄ powder can be improved with better packing with smaller void volumes. High tap density leads to a high volumetric specific capacity. The tap densities of the positive electrode composed of individual LiFePO₄ nanocrystals, dumbbell- and donut-shaped hierarchical microstructures are 1.56, 1.20 and 1.06 g cm⁻³, respectively. As the nanocrystalline LiFePO₄ possesses the highest tap density among the various morphologies, a high volumetric specific capacity (260.5 mAh cm⁻³) was achieved. The cycle performances of the individual nanocrystalline particles, dumbbell-shaped and donut-shaped hierarchical microstructures of LiFePO₄ coated with PEDOT (10 wt%) were tested. The discharge specific capacities for 100 battery cycles are shown in figure 4(B). Specific capacities of 153, 107 and 102 mAh g⁻¹ after 100 charge-discharge cycles at a rate of 1 C, with degradation rated of approximately 10, 11 and 15%, respectively, for the individual nanocrystals, dumbbell-shaped and donut-shaped hierarchical microstructures of LiFePO₄ coated with PEDOT (10 wt%).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Optimization of the amount of PEDOT coated on the LiFePO₄ nanoparticles was required to achieve the long cycle life and the high capacity. Figure 5 shows the battery performances of LiFePO₄ nanocrystals coated with 10, 20 and 30 wt% of PEDOT at a rate of 1 C for the consecutive 100 charge-discharge cycles. As shown in figure 5, the nanocrystals LiFePO₄ coated with PEDOT 20 and 30 wt% exhibited higher specific capacities than the theoretical capacity of LiFePO₄ only; the first discharge capacity of around 190 mAh g⁻¹ for 30 wt% PEDOT coated LiFePO₄ cathode and 175 mAh g⁻¹ for 20 wt% PEDOT coated LiFePO₄ cathode. The capacity loss for the LiFePO₄ coated with 30 wt% of PEDOT, 15% for 100 cycles, was the largest among the three cathodes, especially during the first five cycles. The cycle performance of LiFePO₄ coated with 20 wt% of PEDOT was superior to the cathodes of 30 or 10 wt% PEDOT coating. The capacity remained fairly constant with high capacity values for 100 battery cycles.

Zoom In Zoom Out Reset image size

Long-term cycle stability at the standard 1 C rate and at a higher discharge current rate of 10 C for the individual nanocrystalline LiFePO₄ coated with PEDOT (20 wt%) was investigated and is shown in figure 6. The specific capacity at a rate of 10 C was 136 mAh g⁻¹ for the first battery cycle and 128 mAh g⁻¹ for the 500th cycle, with a degradation rate of approximately 6%. The specific capacity was 163 mAh g⁻¹ after 500 battery cycles at the rate of 1 C, with a degradation rate of 7%. The nanocrystalline LiFePO₄ particles coated with PEDOT (20 wt%) exhibited the high discharge capacities and the long-term cycleability. We believe the mechanical flexibility of the conductive polymer can give room for structural expansion and contraction modification caused by Li insertion and extraction, which results in the long-term cycle stability for the nanocrystalline LiFePO₄ coated with PEDOT cathode.

Zoom In Zoom Out Reset image size