Advancing Particle-Particle Random Phase Approximation: Accurate Double Excitation Prediction and a New Library

The particle-particle random phase approximation (ppRPA) accurately treats two electrons in a CI framework while integrating with DFT description for the remaining (N − 2) electrons. This enables ppRPA to describe correlations in diradical systems and predict singlet-triplet gaps, charge-transfer (CT), and Rydberg excitations, with recent active space developments reducing its computational costs and expanding its applicability. We present two key advancements in ppRPA. First, we demonstrate its accuracy and efficiency in predicting double excitation energies, showing that ppRPA, combined with functionals containing appropriate exact exchange, achieves accuracy comparable to high-level methods like CCSDT and CASPT2 while significantly lowering computational costs. We also introduce a novel application of ppRPA starting from a ΔSCF excited (N − 2)-electron state instead of its ground state used in all previous studies. Additionally, we introduce a newly developed ppRPA library, offering the theoretical chemistry community an accessible tool for a wide range of electronic structure problems. These advancements enhance the utility of ppRPA in molecular and extended systems.