Search

The previously presented mesoscopic model [Phys. Chem. Chem. Phys., 16, 22555, (2014)] for battery electrodes consisting of phase-change insertion materials is incorporated into porous-electrode theory and validated by comparing the simulation results with experimental data from continuous and intermittent galvanostatic discharge of a LiFePO4 electrode under various operating conditions. The model features mesoscopic LiFePO4 units that undergo non-equilibrium lithiation/delithiation and fast solid-state diffusion. Good agreement with the experimental data supports the validity of this model. GITT analysis suggests that the slow evolution of the electrode polarization during each pulse and the subsequent relaxation period is due to Li transport between LiFePO4 units rather than diffusion within the units. Galvanostatic pulse techniques commonly used to determine diffusivities of inserted species in solid-solution systems may also be used to estimate the equilibrium potential of individual mesoscopic units for which no actual measurement has been reported to date. Further analysis of the GITT experiments suggests an alternative pathway for the intermittent charge/discharge of LFP electrodes. Depending on the overall depth-of-discharge/charge of the electrode, relaxation time and the incremental depth-of-discharge/charge of each pulse, the solid-solution capacity available in the Li-rich/Li-poor end-member may be able to accommodate Li insertion/extraction entirely without phase transformation during each pulse. [ABSTRACT FROM AUTHOR]

A 2-D mathematical model of the secondary current distribution in a three-electrode cell is used in order to understand distortion of impedance spectra influenced by the position of the reference electrode (RE) in a cell when one electrode extends past the other. The cell setup consists of a modified coin cell, i.e., working and counter electrodes face each other, and a RE is positioned behind the counter electrode (CE) that has a small hole in it to allow for electrolyte access. This configuration shows large distortion of WE and CE impedance data when taken against the RE. The WE/RE impedance is underestimated and contains a negative impedance contribution from the CE whereas the CE/RE impedance is overestimated and contains a positive impedance contribution from the WE. As supported by simulations, the distortions arise from a radial ionic current flowing from the WE toward the edge of the CE hole, and are minimized through an increase of the electrolyte conductivity, an increase of the separator thickness, or a decrease of the hole size in the CE. The distortions nearly disappear when an additional hole is included in the WE that is perfectly aligned with the hole in the CE. These results should be considered whenever designing a 3-electrode cell where the RE is not sandwiched between the two other electrodes. [ABSTRACT FROM AUTHOR]

The variable solid-state diffusivity (VSSD) and the resistive-reactant (RR) models that focus on different physical phenomena are used to investigate the solid-state transport (bulk effects) and electronic conductivity (surface effects) of LiFeP04 (LFP). For the first time, the models are effectively validated against experimental galvanostatic discharge data over a full range of applied currents. To achieve a reasonable degree of accuracy, particle-level parameters are estimated by fitting to experimental data obtained under low-rate discharge conditions, whereas electrode-level properties are derived based on high-rate conditions. Particle size distribution turns out to play a pivotal role in determining the rate capability of the electrode determined by the VSSD and a revised version of the RR model. Based on the full-range comparative study, both the resistive-reactant effect and bulk-related rate limitations prove to contribute significantly to the electrode polarization, especially at high C-rate. The resistive-reactant effect is expected to increase in an electrode made of smaller LFP particles. [ABSTRACT FROM AUTHOR]

An aging mode! of a commercial graphite/LiFePO4 cell is developed that takes into account side-reaction kinetics and solvent- diffusion limitations across growing passive films at both electrodes. Additionally, a progressive loss of active material at the anode is implemented as an empirical function of current density, temperature and lithium content in the anode particles. The model is calibrated using l-year aging experiments under eight different conditions, consisting either of cycling or storage. Side-reaction and passive-film parameters are adjusted by means of the aging data of OCP-stored cells, and cycling conditions are simulated based on that set of parameters. A good agreement is found between cycling experiments and simulations if the loss of anode active material is dependent on the lithium content in the particles, with more loss experienced for the Li-rich particles. Still, simulations progressively depart from experiments for the most severe conditions (cycling at 45°C), which suggests the existence of an additional aging phenomenon that is not accounted for in the present model. Once the model is validated, simulations are readily performed at longer time than experiments in order to determine the battery lifetime under various aging conditions, providing they remain in the range of applicability of the model. [ABSTRACT FROM AUTHOR]

A mathematical model of the LiFePO4 electrode is developed that is based on a thermodynamically-consistent treatment of solid-state lithium transport, in which the active material is described as a nonideal binary solution of Li-intercalated and empty sites. The nonideality is derived from the experimental open-circuit potential and manifested in a concentration-dependent diffusion coefficient throughout the entire composition range. Under certain operating conditions, a diffuse phase-boundary between Li-poor and Li-rich area is predicted. Furthermore, the actual active-particle-size distribution within the electrodes is captured by three different particle groups in the model. Without embedding porous-electrode effects into the model, simulation/experiment comparisons for three electrodes recovered from different commercial cells at rates up to 1C show the robustness of the variable solid-state diffusivity and particle-size distribution for simulating galvanostatic charges/discharges. In addition, it allows for analysis of the experimental data of various electrodes and to understand their rate limitations. Based on model-parameter comparison between the three designs, the resistive-reactant effect is identified as an additional limiting effect in the electrode comprising nano-particles. The proposed model is a promising candidate for various macroscopic applications, e.g., implementation into 3D battery-pack models for battery management system or comprehensive aging studies. [ABSTRACT FROM AUTHOR]

Capacity loss of a commercial graphite/LiFePO4 cell during either open-circuit-potential storage or under cycling conditions at 25 and 45°C during one year is analyzed with the aid of postmortem analyses and simulations of the cell performance decay over the course of aging. An in-depth understanding of capacity-loss mechanisms under both storage and cycling conditions is gained by refining some parameters of a single-particle model of the cell at different extents of aging. The simulation-based analysis of the aging data reveals that the capacity fade during cell storage only results from the loss of cyclable lithium because of side reactions whereas the loss of graphite active material is an additional source of aging for the cells under cycling conditions. A simple kinetic analysis of electrode/electrolyte interactions is provided for the cells under storage conditions. Moreover, the growth of solid-electrolyte interphase (SEI) at the graphite electrode under storage conditions is simulated in order to refine the solvent-reduction kinetic parameters and solvent diffusion coefficient in the SEI layer. From the analysis, it is shown that the SEI growth during storage is under mixed kinetic/diffusion control. [ABSTRACT FROM AUTHOR]

Lithium diffusion coefficient (D) in LiyFeSO4F is investigated by means of three electrochemical techniques, namely potentiostatic intermittent titration technique (PITT), galvanostatic intermittent titration technique (GITT), and galvanostatic charge/discharge at different C-rates. The analytic equations used for analyzing PITT and GITT are based on radial diffusion in spherical particles, which is more suitable to the type of electrode studied here than the original equations derived by Wen et al. [J. Electrochem. Soc., 126, 2258 (1979).] for a slab geometry. While PITT and GITT analyses with the analytic equations are only valid in the partial solid solution regions (y ≤0.29 and y ≤0.83 in LiyFeSO4F), the analysis of galvanostatic charge/discharge with a mathematical model allows for the determination of D over the entire lithium composition range. The average D values found with PITT and GITT analyses are in good agreement with each other, and of the same order as those reported for LiyFePO4 in the literature (ca. 10-14 cm2/s); However, D is larger for LiyFeSO4F in the Li-rich composition range than in the Li-poor one, whereas it is the opposite for LiyFePO4. This result could explain why LiyFeSO4F has a better rate capability on discharge than LiyFePO4. [ABSTRACT FROM AUTHOR]

A methodology for predicting the capacity loss of a battery under a complex current profile is provided. The prediction is based upon a set of elementary aging experiments, which are herein generated by a physics-based model featuring a surrogate battery, for which the source of aging is the growth of a solid-electrolyte interphase at the anode. Empirical correlations of the capacity loss as a function of the aging time and of the aging time as a function of the capacity loss are developed based on 12 dummy elementary aging experiments under different conditions of state of charge and current. Those correlations are used together with the relationship for loss accumulation over time or the Palmgren-Miner rule (both introduced in Part I) to predict the aging of a complex current profile. The prediction accuracy is assessed by a direct simulation of the complex profile by using the physics-based model featuring the surrogate battery. [ABSTRACT FROM AUTHOR]

Li-ion batteries have contributed to the commercial success of portable electronics, and are now in a position to influence higher-volume applications such as plug-in hybrid electric vehicles. Most commercial Li-ion batteries use positive electrodes based on lithium cobalt oxides. Despite showing a lower voltage than cobalt-based systems (3.45 V versus 4 V) and a lower energy density, LiFePO4 has emerged as a promising contender owing to the cost sensitivity of higher-volume markets. LiFePO4 also shows intrinsically low ionic and electronic transport, necessitating nanosizing and/or carbon coating. Clearly, there is a need for inexpensive materials with higher energy densities. Although this could in principle be achieved by introducing fluorine and by replacing phosphate groups with more electron-withdrawing sulphate groups, this avenue has remained unexplored. Herein, we synthesize and show promising electrode performance for LiFeSO4F. This material shows a slightly higher voltage (3.6 V versus Li) than LiFePO4 and suppresses the need for nanosizing or carbon coating while sharing the same cost advantage. This work not only provides a positive-electrode contender to rival LiFePO4, but also suggests that broad classes of fluoro-oxyanion materials could be discovered. [ABSTRACT FROM AUTHOR]

Abstract: A study of LiFePO4-based electrodes prepared through various synthesis conditions is presented. From X-Ray diffraction, high resolution transmission electron microscopy, electrochemical Li+ extraction/insertion and electrical conductivity data we conclude that the use of starting precursors such as Li2CO3, FeC2O4·2H2O and/or Nb(OC6H5)5 produces LiFePO4-based composites containing significant amounts of carbon. We never succeeded in doping LiFePO4 with Nb to yield Li1−x Nb x FePO4 but produced, instead, crystalline β-NbOPO4 and/or an amorphous (Nb, Fe, C, O, P) “cobweb” around LiFePO4 particles which is responsible for superior electrochemical activity. AC-conductivity measurements conclude to a total electrical conductivity of ∼10−9 S cm−1 at 25 °C with an activation energy of ca. 0.65 eV for pure LiFePO4 and LiFePO4/β-NbOPO4 composites. C-containing LiFePO4 samples, including those that were tentatively but unsuccessfully doped with Nb, are much more conductive (up to 1.6·10−1 S cm−1) with an activation energy ΔE∼0.08 eV. [Copyright &y& Elsevier]