Efficient and reliable modeling of large $π$-electron systems with the Pariser--Parr--Pople Hamiltonian and pCCD-based methods

Abstract

Model Hamiltonians offer a cost-effective way to capture the key physics of large $π$-conjugated systems. In this work, we combine the Pariser--Parr--Pople (PPP) model Hamiltonian with pair Coupled Cluster Doubles (pCCD)-based methods to study the excited-state electronic structures of polycyclic aromatic hydrocarbons (PAHs). The model Hamiltonian is implemented in the open-source PyBEST software package, which provides numerous pCCD-type models that have been shown to perform well for the electronic structure properties of large organic molecules when combined with quantum-chemical Hamiltonians and all-electron basis sets. Within the PPP model, we probe canonical Hartree--Fock and natural pCCD-optimized orbitals in predicting excited-state properties using the linear response formalism on top of pCCD. Their performance is compared with configuration-interaction-based methods and the conventional EOM-CCSD approach. Our study demonstrates that pCCD/PPP-based approaches are an efficient, cost-effective, and accurate framework to compute excited-state properties in large $π$-conjugated organic systems, such as extended PAHs or other systems relevant to organic electronics.