The performance of lithium-ion batteries (LIB) is strongly influenced by the composition and fabrication of the electrode structures. On the one hand, the used material and its microstructure plays an important role. On the other hand, the mechanical densification processes impact on the quality of the battery. Here, the effective conductivities of both the solid phase and the electrolyte phase are being considered as the performance parameters and are being modelled using discrete element methods.
In order to study the influence of the microstructure on the performance and to come up with an electrode design, we use a certain toolbox to first generate initial microstructures, densify those and calculate the effective conductivities. To generate an initial microstructure, we use the Random-Close-Packing algorithm (RCP), where a randomly distributed, densely packed and overlap-free assembly of spheres can be created. After the densification processes, where the Discrete-Element-Method (DEM) or the Numerical-Sintering algorithm (NS) is being used, we can calculate effective conductivities. To that end, we convert the transport participants of both the electronic and the ionic conductivity into an equivalent circuit of nodes and resistors, i.e. the Resistor-Network (RN). Using a combination of Kirchhoff’s and Ohm’s law, the RN becomes solvable for the unknown current and therefore the effective conductivity can be calculated.
There are several ways one can enhance the conductivity of a cathode. For one, due to the low electrical conductivity of the active material (AM) inside a LIB electrode, carbon black (CB) powder is being added to overcome this drawback. Further, the electrode is being densified mechanically to establish more conducting pathways and therefore enhancing conductivity. However, this method leads to a decrease in the electrolyte volume or the pore space, respectively, which results in a lower ionic conductivity. Lastly, the AM can be coated with a thin film of carbon, which again increases the electronic conductivity.
In the presented work, discrete element methods are being presented to model both the electronic and ionic conductivity. Due to the efficiency of those methods concerning the calculation time and computational resources, parametric studies are made possible. Thus, optimized microstructures with any given material can be found, such that electronic and ionic conductivity would reach the best value.