No abstract

LiFePO 4 /Li 4/3 Ti 5/3 O 4 Li-ion cells have been investigated by many groups and their behavior in standard electrolytes such as 1 M LiPF 6 ethylene carbonate: diethyl carbonate ͑EC:DEC͒ is well known. Here we report on the behavior of these cells with 2,5-ditertbutyl-1,4-dimethoxybenzene added to the electrolyte as a redox shuttle additive to prevent overcharge and overdischarge. We explore methods to increase the current-carrying capacity of the shuttle and explore the heating of practical cells during extended overcharge. The solubility of 2,5-ditertbutyl-1,4-dimethoxybenzene was determined as a function of salt concentration in lithium bis-oxolatoborate-͑LiBOB͒ and LiPF 6 -containing electrolytes based on propylene carbonate ͑PC͒, EC, DEC, and dimethyl carbonate ͑DMC͒ solvents. Concentrations of 2,5-ditertbutyl-1,4-dimethoxybenzene up to 0.4 M can be obtained in 0.5 M LiBOB PC:DEC ͑1:2 by volume͒. Coin-type test cells were tested for extended overcharge protection using an electrolyte containing 0.2 M 2,5-ditertbutyl-1,4-dimethoxybenzene in 0.5 M LiBOB PC:DEC. Sustained overcharge protection at a current density of 2.3 mA/cm 2 was possible and hundreds of 100% shuttle-protected overcharge cycles were achieved at current densities of about 1 mA/cm 2 . The diffusion coefficient of the shuttle molecule in this electrolyte was determined to be 1.6 ϫ 10 −6 cm 2 /s from cyclic voltammetry and also from measurements of the shuttle potential vs. current density. The power produced during overcharge was measured using isothermal microcalorimetry and found to be IV as expected, where I is the charging current and V is the cell terminal voltage during shuttle-protected overcharge. Calculations of the temperature of 18650-sized Li-ion cells as a function of time during extended shuttle-protected overcharge at various C-rates are presented. These show that Li-ion cells need external cooling during extended shuttle-protected overcharge if currents exceed about C/5 rates.

Fifty eight aromatic organic molecules were screened as chemical shuttles to provide overcharge protection for LiFePnormalO4 /graphite and LiFePnormalO4∕normalLi4∕3normalTi5∕3normalO4 Li-ion cells. The majority of the molecules were based on methoxybenzene and on dimethoxybenzene with a variety of ligands added to explore their effect. The added ligands affect the redox potential of the molecules through their electron-withdrawing effect and affect the stability of the radical cation. Of all the molecules tested, only 2,5-di-tert-butyl-1,4-dimethoxybenzene shows an appropriate redox potential of 3.9V vs Li∕normalLi+ and long-term stability during extended abusive overcharge totaling over 300cycles of 100% overcharge per cycle. The reasons for the success of this molecule are explored.

The oxidation potentials of seventeen molecules used as candidate shuttle additives in Li-ion cells were calculated using density functional theory and compared with experiment. The root-mean-square deviation between the calculated and measured oxidation potentials of these seventeen molecules is 0.15 V with the maximum deviation being 0.25 V, indicating that the ab initio calculation is in good agreement with the experiment. Neglecting thermal contributions in the calculation of standard oxidation potentials at ambient conditions does not lead to significant errors. An empirical relation between the oxidation potentials and the orbital energies of these molecules in solution is presented. The oxidation potential of a molecule could be estimated based on the orbital energy of the molecule's highest occupied molecular orbital or its cation's lowest unoccupied molecular orbital in solution with an error less than 0.3 V for most molecules reported.Redox shuttle electrolyte additives in lithium and lithium-ion batteries have been long proposed to protect against the overcharge of cells in series-connected batteries. 1-10 The shuttle molecule has a defined oxidation potential at which it gives an electron to the positive electrodeIt then travels to the negative electrode and the reverse reaction occursThe electron travels between the electrodes in the external circuit. The shuttle molecule S then diffuses back to the positive electrode where Reaction 1 takes place again. The shuttle molecule therefore carries the charging current at a defined potential. In a recent paper, 11 we reported the screening of 58 aromatic organic molecules as chemical shuttle candidates to provide overcharge protection for LiFePO 4 /graphite and LiFePO 4 /Li 4/3 Ti 5/3 O 4 Li-ion cells. It resulted in the successful identification of 2,5-di-tert-butyl-1,4-dimethoxybenzene as a redox shuttle additive to protect against both the overcharge and overdischarge of LiFePO 4 -based Li-ion cells. 12 Meanwhile, in cooperation with the experiment we have systematically performed quantum chemical calculations on the candidate shuttle molecules mainly to evaluate their oxidation potentials. In this paper, we report computational and experimental results for the oxidation potentials for 17 molecules covering four classes of molecules/radicals: 8 aromatic molecules, 3 TEMPO or 2,2,6,6-tetramethylpiperidinyloxy-like radicals, 3 pyridine NO-like molecules, and 3 N-substituted phenothiazine molecules. The agreement between the calculation and experiment is very good. The root-mean-square deviation between the calculated and measured oxidation potentials of these 17 molecules is 0.15 V, with the maximum deviation being 0.25 V. With such a precision, quantum chemical calculation becomes a powerful tool in the processes of searching for new chemical shuttle candidates. We may eliminate those candidates for which the calculated oxidation potentials are out of the desired range. By modifying existing molecules, the redox potentials can be estimated through calcula...

The molecules 10-methylphenothiazine, 10-ethylphenothiazine, 3-chloro-10-methylphenothiazine, 10-isopropylphenothiazine, and 10-acetylphenothiazine are shown to be stable redox shuttle additives in LiFePO 4 /graphite and LiFePO 4 /Li 4/3 Ti 5/3 O 4 Li-ion coin cells to protect against overcharge and overdischarge. The diffusion constant of 10-methylphenothiazine was measured using cyclic voltammetry to be 1.5 ϫ 10 −6 cm 2 /s, which translates to maximum shuttle-protected overcharge current densities near 2 mA/cm 2 in practical cells. Although the redox potentials of these molecules ͑near 3.5 V͒ are somewhat low for LiFePO 4 , their stability over repeated overcharge and overdischarge cycles is about equal to that of 2,5-di-tert-butyl-1,4-dimethoxybenzene that has been shown to provide protection for over 200 cycles of 100% overcharge at C/10. Given the stability of these oxidized molecules, we believe that phenothiazine represents an attractive core for ligand substitution to adjust the redox potential to more practical values.