Doctor Stephen J. Harris
Lawrence Berkeley National Laboratory
A long-term goal of DOE is the development of batteries that could power a commercially successful all-electric vehicle, probably requiring a range near 200 miles (at least double that of a Nissan Leaf). In order to achieve higher energy densities, current research focuses on new electrode chemistries with higher capacities and higher voltages (volumetric energy density = charge capacity voltage mass density) that could provide a >50% increase in energy density—still insufficient for a 200 mile range. (Volumetric energy density is more critical to auto companies than gravimetric energy density.) Improvement in the mass density term has been stymied because low electrode porosities always seem to lead to high tortuosities and low power densities. We note, however, that the inverse relationship between porosity and tortuosity applies only to electrodes with random microstructures, which inevitably have large local inhomogeneities.
At the same time, failure in materials almost always begins at local inhomogeneities. Yet, current analyses of lithium-ion batteries are based on a porous electrode model that assumes a homogeneous mixture of flawless, isotropic particles. Our work suggests that variability in local microstructure and in internal particle morphology plays a critical role in reducing both performance and durability of Li-ion batteries. Prof. Edwin Garcia’s models lead to similar conclusions.
Following this reasoning, and in collaboration with Garcia, we are developing a new approach for electrode architectures. We suggest that with designed (i.e., non-random) microstructures and particles, we can create electrodes with a high density of active material, to maximize capacity; a low tortuosity, to maximize power; and improved uniformity (no weak spots), to maximize battery life.