Publications

A new varying current decay (VCD) protocol, which charges the Li-ion battery at a faster rate, was developed. The performance of the battery charged using the VCD protocol was compared with the performance of batteries charged with conventional constant current-constant voltage (CC-CV) and constant voltage (CV) protocols. The destructive physical analysis tests at the end of 150 cycles indicated higher impedance for the cells cycled using the VCD protocol compared to the cell charged using the conventional (CC-CV) mode. The observed increase of the impedance was due to a small increase of the potential above the cut-off value of 4.2 for short times. A complete capacity fade material balance as a function of number of cycles was performed in order to account for the total capacity loss due to different charging protocols used. The loss of primary active material (Li+), the secondary active material (LiCoO2/carbon) and the rate capability losses were determined for Sony US 18650 Li-ion cells and compared for different charging protocols. © 2003 Elsevier Science B.V. All rights reserved.

In this study a three-phase homogeneous model was developed to simulate the performance of the molten carbonate fuel cell (MCFC) cathode. The homogeneous model is based on volume averaging of different variables in the three phases over a small volume element. This approach can be used to model porous electrodes as it represents the real system much better than the conventional agglomerate model. Using the homogeneous model the polarization characteristics of the MCFC cathode was studied under different operating conditions.

A characterization of the ionic conduction of the active layer of a polymer electrolyte membrane fuel cell (PEMFC) cathode by ac impedance measurement at open-circuit potential conditions was conducted. Porous electrode theory was used to derive a compact equation, partial derivative(2)(&UPhi;) over cap (2)/partial derivativey(2) + partial derivative ln f(y)/partial derivativey x partial derivative(&UPhi;) over cap (2)/partial derivativey - R/f (y)(1 + jOmega)(&UPhi;) over cap (2) = 0, to solve for the impedance response of a cathode at open-circuit potential conditions. This equation includes a parameter R, the ratio of an ionic resistance (evaluated at the active layer/membrane interface), to the total charge-transfer resistance of the active layer. The influence of an assumed ionic conductivity distribution profile f ( y) on the error in the estimation of total double-layer capacitance of the active layer from the -1/(Z(Im)omega) vs. Z(Re) plot was also investigated in this work. The increase of ionic conductivity in the active layer of an air cathode with an increase in the ionomer loading was revealed from both impedance data and surface area measurements. A nonlinear parameter estimation method was used to extract the ionic resistance from the high-frequency region of the impedance data at open-circuit potential conditions. The assumed ionic conductivity distribution profile in the active layer was found to vary with ionomer loadings. (C) 2003 The Electrochemical Society.

A complete capacity fade analysis was carried out for Sony 18650 cells cycled at elevated temperatures. The major causes of capacity loss were identified and a complete capacity fade balance was carried out to account for the total capacity loss of Li-ion battery as a function of cycle number and temperature. The three most significant parameters that cause capacity loss were loss of secondary active material (LiCoO2/carbon) and primary active material (Li+) and the rate capability losses. Intrinsic capacity measurements for both positive and negative electrode has been used to estimate the capacity loss due to secondary active material and a charge balance gives the capacity lost due to primary active material (Li+). Capacity fade has been quantified with secondary active material loss dominating the other losses. © 2002 Elsevier Science B.V. All rights reserved.

A review of mathematical models of lithium and nickel battery systems developed at the University of South Carolina is presented. Models of Li/Li-ion batteries are reviewed that simulated the behavior of single electrode particles, single electrodes, full cells and batteries (sets of full cells) under a variety of operating conditions (e.g. constant current discharge, pulse discharge, impedance and cyclic voltammetry). Models of nickel battery systems are reviewed that simulate the performance of full cells, as well as the behavior of the nickel hydroxide active material. The ability of these models to predict reality is demonstrated by frequent comparisons with experimental data. © 2002 Elsevier Science B.V. All rights reserved.

This work presents a rigorous continuum model describing the transport of ions and assocd. ion pairs in solid polymer electrolytes subjected to small amplitude a.c. (ac) excitation. The model treats ion assocn. as a reversible reaction among ions and ion pairs. Dimensionless governing equations are developed from component mass balances, flux equations based on dil. soln. theory, and the Poisson equation. Assuming reversible electrode reactions and electroneutrality, the model equations have an anal. soln. Further simplifications are possible in limiting cases (weak and strong assocn., zero and infinite frequency excitation), giving expressions consistent with previously published models. We use the model to explore the effect of assocn./dissocn. reaction rates, ion pair diffusivity, and fractional dissocn. on ac impedance behavior. We present a scheme for establishing component diffusivities and fractional dissocn. from independent exptl. data for lithium perchlorate in poly(ethylene oxide). With no addnl. adjusted parameters, satisfactory agreement exists between calcd. and exptl. ac impedance data

The validity of estimating the solid phase diffusion coefficient, Ds, of a lithium intercalation electrode from impedance measurement by a modified electrochemical impedance spectroscopy (EIS) method is studied. A macroscopic porous electrode model and concentrated electrolyte theory are used to simulate the synthetic impedance data. The modified EIS method is applied for estimating Ds. The influence of parameters such as the exchange current density, radius of active material particle, solid phase conductive, porosity, volume fraction of inert material, and thickness of the porous carbon intercalation electrode, the solution phase diffusion coefficient, and transference number, on the validity of Ds estimation, is evaluated. A simple dimensionless group is developed to correlate all the results. It shows that the accurate estimation of Ds requires large particle size, small electrode thickness, large solution diffusion coefficient, and low active material loading. Finally, a "full model" method is developed for the cases where the modified EIS method does not work well.

The performance of Cell-Batt® Li-ion cells and Sony 18650 cells using non-stoichiometric spinel and LiCoO2, respectively, as positive electrode material has been studied under several modes of charging. During cycling, the cells were opened at intermittent cycles and extensive material and electrochemical characterization was done on the active material at both electrodes. Capacity fade of spinel-based Li-ion cells was attributed to structural degradation at the cathode and loss of active material at both electrodes due to electrolyte oxidation. For the Sony cells both primary (Li+) and secondary active material (LiCoO2)/C) are lost during cycling. © 2002 Elsevier Science B.V. All rights reserved.

Cobalt substituted lithium-nickel oxides were synthesized by a solid-state reaction procedure using lithium nitrate, nickel hydroxide and cobalt oxalate precursor and were characterized as cathodes for molten carbonate fuel cell (MCFC). LiNi0.8Co0.2O2 cathodes were prepared using non-aqueous tape casting technique followed by sintering in air. The X-ray diffraction (XRD) analysis of sintered LiNi1-xCoxO2 indicated that lithium evaporation occurs during heating. The lithium loss decreases with an increase of the cobalt content in the mixed oxides. The stability studies showed that dissolution of nickel into the molten carbonate melt is smaller in the case of LiNi1-xCoxO2 cathodes compared to the dissolution values reported in the literature for state-of-the-art NiO. Pore volume analysis of the sintered electrode indicated a mean pore size of 3 $μ$m and a porosity of 40%. A current density of 160 mA/cm2 was observed when LiNi0.8Co0.2O2 cathodes were polarized at 140 mV. The electrochemical impedance spectroscopy (EIS) studies done on LiNi0.8Co0.2O2 cathodes under different gas conditions indicated that the rate of the cathodic discharge reaction depends on the O2 and CO2 partial pressures. © 2002 Published by Elsevier Science B.V.