Accurate and robust wavefunction methods for energies andstructures of electronically excites states in complex systems

点击翻译按钮获取中文摘要

Methods based on second-order perturbation theory as the approximate coupled-cluster singles-and-doubles model CC2 are widely applied for the prediction of UV/Vis spectra and potential energy surfaces for excited states. They are often the only tractable first principles approaches if time-dependent density functional theory is not applicable because charge-transfer processes are involved. For systems beyond 100 atoms, the use of these methods is hampered by their steep cost-scaling.Another limitation are large errors for doubly excited states and if strong correlation effects are encountered in the ground state. These problems can very often be overcome with the full coupled-cluster singles-and-doubles (CCSD) and the approximate triples model CC3, which have, however, an even steeper cost-scaling.In this project we will develop an implementation of the hierarchy CC2, CCSD, CC3 and approximations to these with reduced computational costs and apply them for systems relevant for photoelectrochemical cells, which are challenging for currently available methods. To achieve a robust accuracy we only apply approximations based on the analysis of wavefunction and energy diagnostics.