Abstract
Optimizing synthesis parameters is the key to successfully design ideal Ni-rich cathode materials that satisfy principal electrochemical specifications. We herein implement machine learning algorithms using 330 experimental datasets, obtained from a controlled environment for reliability, to construct a predictive model. First, correlation values showed that the calcination temperature and the size of the particles are determining factors for achieving a long cycle life. Then, we compared the accuracy of seven different machine learning algorithms for predicting the initial capacity, capacity retention rate, and amount of residual Li. Remarkable predictive capability was obtained with the average value of coefficient of determinant, R(2) = 0.833, from the extremely randomized tree with adaptive boosting algorithm. Furthermore, we propose a reverse engineering framework to search for experimental parameters that satisfy the target electrochemical specification. The proposed results were validated by experiments. The current results demonstrate that machine learning has great potential to accelerate the optimization process for the commercialization of cathode materials.