Bond sensitive graph neural networks for predicting high temperature superconductors

Finding high temperature superconductors (HTS) has been a continuing challenge due to the difficulty in predicting the transition temperature (T_c) of superconductors. Recently, the efficiency of predicting T_c has been greatly improved via machine learning (ML). Unfortunately, prevailing ML models have not shown adequate generalization ability to find new HTS, yet. In this work, a graph neural network model is trained to predict the maximal T_c (T_c^max) of various materials. Our model reveals a close connection between T_c^max and chemical bonds. It suggests that shorter bond lengths are favored by high T_c, which is in coherence with previous domain knowledge. More importantly, it also indicates that chemical bonds consisting of some specific chemical elements are responsible for high T_c, which is new even to the human experts. It can provide a convenient guidance to the materials scientists in search of HTS.