A user-friendly differential voltage analysis software has been developed and is described here. High-precision reference potentialspecific capacity data for Li/negative electrode and Li/positive electrodes, as well as the cycled full cell potential-specific capacity, must be supplied by the user. From these, the differential voltage versus capacity, dV/dQ vs. Q, of a full Li-ion cell is calculated and compared to experiment. The calculated dV/dQ vs. Q curve has four adjustable parameters which are optimized manually with slider bars or automatically by least squares fitting of the calculation to experiment. The parameters are the positive electrode mass, the negative electrode mass, the positive electrode slippage and the negative electrode slippage. Examples of the use of the program are given for graphite/LiCoO 2 wound cells cycled for hundreds of cycles. The variation of the four parameters with cycle number give insights into the mechanisms of cell failure equivalent to that which could be obtained with a Li reference electrode inserted within the cell. The software is available free of charge by contacting the authors.

The electrochemical performance of negative electrodes based on commercially available micrometer-sized ␣-Fe 2 O 3 powder and four different binders was investigated. ␣-Fe 2 O 3 electrodes made using sodium carboxymethyl cellulose binder and two proprietary binders show better cycling performance than electrodes made from the conventional binder, polyvinylidene fluoride ͑PVDF͒. Heat-treating the PVDF electrodes to 300°C improves electrode performance dramatically. A specific capacity of over 800 mAh/g for over 100 cycles has been achieved for Li/␣-Fe 2 O 3 cells with micrometer-sized ␣-Fe 2 O 3 , by contrast to many teachings in the literature where it is claimed nanometer sized particles are required to obtain such performance. As is the case for composite electrodes made from powders of metal alloys, we suggest that binder choice has a great impact on the performance of metal oxide electrodes demonstrating large volume expansion during lithiation, such as ␣-Fe 2 O 3 .

In order to develop Li-ion batteries with improved lifetimes, a means of quickly and accurately estimating battery life is required. The use of coulombic efficiency (CE) is an important tool which provides a way to quantify parasitic reactions occurring within cells. As more stable battery chemistries are developed, the rates of parasitic reactions become smaller and differences in CE among cells with different electrolyte additives become increasingly smaller. In order to resolve these differences, charger systems must be developed which can measure CE with increased precision and accuracy. This paper investigates various ways to improve the precision and accuracy of CE measurements. Using the high-precision charger (HPC) at Dalhousie University (built in 2009) as a starting point, a new prototype charger was built with several modifications to the design of the existing HPC. The effect of each of these modifications is investigated in detail to provide a blueprint for the development of next-generation charger systems. This prototype charger shows greatly improved precision and accuracy, with CE results that are approximately four times more precise than those of the existing HPC and over an order of magnitude more precise than high-end commercially available charger systems.

The capacity loss and coulombic efficiency of commercial LiCoO 2 /graphite Li-ion cells have been examined using high precision coulometry and long-term cycling tests. The experiments show that time, not cycle count, was the dominant contributor to the degradation of LiCoO 2 /graphite Li-ion cells cycled at low rates and elevated temperatures. The differential voltage versus capacity, dV/dQ vs Q, of the cells was measured for all cycles during the extended cycling and fit using predicted dV/dQ vs Q plots calculated from Li/graphite and Li/LiCoO 2 cells made from the same electrodes that were used in the commercial cells. From this analysis it was possible to determine fraction of positive and negative electrode masses that remained active as a function of cycle number and also the portion of capacity loss due to the relative slippage of the positive and negative electrode potential-capacity curves. A rapid impedance rise was observed near the end of the cycling testing. These results provide a model procedure for understanding of the failure of lithium-ion cells subjected to sustained high temperature cycling.

Active lithium in the negative electrode of a Li-ion cell reacts with electrolyte to form an ever-thickening solid electrolyte interphase. The rate of this reaction can be monitored as a function of temperature, time and electrode potential using storage and symmetric cell studies. Using the soft carbon, petroleum coke, as a model negative electrode material, experiments measuring the open circuit voltage (OCV) change with time of Li/coke cells were made to measure the rate of loss of active lithium. The capacity loss with cycle number or time of coke/coke symmetric cells was also used to measure the rate of Li loss. The results on over 100 test cells show that: 1) the reaction rate decreases by about a factor of 2-4 as the electrode potential increases from 0.005 to 1.0 V; 2) the reaction rate increases approximately 3-10 fold between 30 and 60 • C depending on the electrode potential; 3) The reaction rates are within a factor of two, and may be the same, for electrodes at OCV or undergoing cycling; 4) the reaction rate is larger when vinylene carbonate (VC) is present in the electrolyte at 30 • C for all potentials and times studied and 5) the reaction rate is about two times smaller in the presence of VC at 60 • C for potentials above 0.4 V. A significant number of further experiments are required to develop accurate theoretical models of the reactivity of intercalated lithium with electrolyte as a function of time, temperature and potential.