Find in Library

Quantum mechanical calculations of molecular ionized states are computationally quite expensive. This work reports a successful extension of a previous deep-neural networks approach towards transferable neural-network models for predicting multiple properties of open shell anions and cations.