167
Cyberdesign Computational Algorithm: Selecting Anode and Cathode Materials through Simulation for Lithium-Ion Batteries Based on Diffusivity

Thursday, 17 May 2018: 08:40
Room 607 (Washington State Convention Center)
Z. Miller, I. Miske, and H. Hosseinzadeh (Rowan University)
A cyber-enabled approach to lithium ion battery design was developed in this work where the user inputs a charging voltage (V), desired charging time (t) in seconds, and desired capacity of charge at time t in mAh/cm2 for their battery. This developed program utilizes both a database brain and computational brain. The computational brain has a computational algorithm for macro/mesoscale Fickian approach of ionic diffusion. This algorithm has been equipped with the Butler-Volmer equation, stochastic methods, Poisson equation solver and other related transport equations. The database brain is at early stage in this software and it has been developed by experimental and computational data. An additional algorithm connects the database and computational brain to the user interface. By submitting a request to the software, the program outputs a combination of anode and cathode materials from an established database or computational brain that could achieve the desired input parameters. At the current state, the program will output a combination of materials from the database that can achieve the desired input parameters if the capacity at time t solved for by the computational brain is equal to the desired capacity established by the user. Further steps are being taken to improve the size of the database, consider systems where the user's parameters are not achievable by materials in the database, and consider other variables in the system besides diffusivity of anode and cathode materials. As an example, the capacity of 200 mAh/cm2 when 5V are applied for 10 seconds was submitted to the software. The software utilized the database and computational brain and it found that a pan-based carbon fiber anode combined with a LixCoO2 cathode charges to 53.823 mAh/cm2 according to simulation when 5V are applied for 10 seconds, which is far from the submitted capacity. Then the software used the computational brain to change the anode material to graphite under the same charging conditions which outputs the same capacity. Then by changing the cathode material to have a diffusivity coefficient one order higher than that of LixCoO2 under the same charging conditions the program was able to output capacities of 221.091 mAh/cm2 or higher.