Geometry Optimization of Porous Electrode for Lithium-Ion Batteries

Abstract

Since the invention of lithium-ion batteries about 40 years ago, the electrodes have been manufactured in essentially the same way: granulating active materials into powder and pasting them on conductive substrate with some additives such as carbon black and binder. Most optimizations have been focusing on tuning parameters such as particle sizes, porosity and electrode thickness, with the assumption of a uniform electrode. Recently some work demonstrated better cell performance by engineering the shape of the electrode with features such as nano cylinders, tunnels and fractal branches. On the simulation side, however, to the best of our knowledge, there has been no report on freeform electrode shape optimization: all current shape optimizations assume a particular electrode pattern with a few adjustable dimensional parameters, and optimize those parameters. Here we report a method to optimize the topology of the porous electrode without a prior assumption of the pattern. An NMC 333 electrode with lithium metal as the counter electrode is considered as an example. To find the optimal solid material distribution for maximum specific energy, we represented the NMC material distribution by 5-by-5 nodal design variables. To solve this 25-dimensional optimization problem, we integrated Deep Neural Network (DNN) into Generalized Simulated Annealing (GSA). During the optimization process, the DNN learns the objective function in real time with training data generated from finite element calculations. GSA uses the DNN to iteratively calculate the optimal solution and generates new training samples based on current optimal geometry to provide feedback to better train the network for further refinement. We show that the method can produce an electrode pattern with the specific energy about 18% higher than that of the uniform electrode. This generic geometry optimization approach provides a powerful tool for the design and optimization of porous electrodes. Figure 1