Mixed-Carbon Anodes for Improved Li-Ion Cells

NASA's Jet Propulsion Laboratory, Pasadena, California

Rechargeable lithium-ion electrochemical cells that contain anodes made from a mixture of graphite and carbon black have been found to perform better than do similar cells that contain anodes made of graphite alone. As explained in more detail below, the addition of carbon black improves performance by increasing effective electrical conductivity.

Typically, the anodes in state-of-the-art lithium-ion cells are made of amorphous carbon (coke) or graphite. Heretofore, these forms of carbon have been considered to have excellent electrical conductivity; therefore, until now, no attempt was made to incorporate carbon black or other conductive diluents into the anodes. Now, however, it has been observed that when a carbon black (more specifically, Shawanigan black) is incorporated into graphite electrodes, cycle-life performance is significantly improved.

Specific Discharge Capacities of four cells were measured in cyclic charging at a current of 30 mA and discharging at a current of 60 mA.

Four lithium-ion cells were fabricated for experiments to determine the effects of incorporating carbon black into graphite anodes. In each cell, the cathode was made of LiCoO₂ and the electrolyte was made of a 1.0 M solution of LiPF₆ in a solvent that consisted of equal volume parts of ethylene carbonate (EC), diethyl carbonate (DEC), and dimethyl carbonate (DMC). The anodes of the four cells had the following compositions:

A commercial graphite (10-28 MCMB);
A mixture of 95 weight percent 10-28 MCMB and 5 weight percent carbon black;
A mixture of 90 weight percent 10-28 MCMB and 10 weight percent carbon black; and
A mixture of 50 weight percent 10-28 MCMB and 50 weight percent of another commercial graphite (KS-44).

In cyclic charge/discharge measurements, all three cells containing mixed-carbon anodes (cells 2, 3, and 4) retained greater proportions of their initial discharge capacities than did the cell containing the single-carbon anode (cell 1). Moreover, at about 15 cycles, the absolute discharge capacity of the single-carbon cell fell significantly below that of all three mixed-carbon cells (see figure).

The tentative explanation for some of these experimental results is the following:

The carbon black particles are significantly smaller than those of 10-28 MCMB. Therefore, in the 10-28 MCMB/carbon-black electrodes, the carbon black particles are conjectured to occupy the voids between the larger 10-28 MCMB particles. In so doing, the carbon black particles increase the effective electrical conductivity of the anode by contributing to overall electrical contact among the anode particles throughout the anode.
The 10-28 MCMB and KS-44 graphites have different particle sizes; consequently, electrical contact and electrical conductivity in the mixed-graphite anode are increased in the same manner as in the 10-28 MCMB/carbon-black electrodes.

While carbon black contributes to electrical conductivity, it also irreversibly consumes some lithium and thereby contributes to a partial irreversible loss of capacity. Apparently, the 5-weight-percent carbon black content provided sufficient electrical contact for increased retention of capacity while contributing little irreversible loss during initial charging. The anode containing 10 weight percent of carbon black exhibited approximately the same retention of capacity, but the initial irreversible loss was considerably greater than in the anode containing 5 weight percent of carbon black. From these observations, one can conclude that 5 weight percent is the optimum amount of carbon black.

This work was done by Chen-Kuo Huang and Jeffrey Sakamoto of Caltech forNASA's Jet Propulsion Laboratory. NPO-20603