An Improved Virtual Orbital Driven Similarity Renormalization Group Approach for Core-Ionized and Core-Excited States

Area of Science:

• Quantum chemistry
• Computational physics
• Spectroscopy

Background:

• Accurate modeling of core-ionized and core-excited states is crucial for understanding molecular electronic structure and chemical dynamics.
• Traditional methods often require computationally expensive orbital optimizations, limiting their applicability to larger systems.

Purpose of the Study:

• To develop and validate a cost-effective computational method for studying core-level electronic states.
• To assess the accuracy of the improved virtual orbitals-generalized active space configuration interaction-driven similarity renormalization group (IVO-GASCI-DSRG) method across various molecular systems.

Main Methods:

• Combining multireference driven similarity renormalization group (DSRG) with improved virtual orbitals (IVOs) and generalized active space configuration interaction (GASCI).
• Testing the IVO-GASCI-DSRG method with three truncation levels on diverse molecular datasets.
• Utilizing third-order perturbative corrections (MRPT3) for enhanced accuracy.

Main Results:

• The IVO-GASCI[3]-DSRG-MRPT3 method demonstrates a favorable balance between computational cost and accuracy.
• The method accurately predicts potential energy surfaces for core-excited and core-ionized states in small molecules (CO, N2, HF).
• Simulations of X-ray absorption spectra for thymine and adenine successfully reproduce key experimental features.

Conclusions:

• The IVO-GASCI-DSRG-MRPT3 method provides a robust and efficient approach for investigating core-level electronic phenomena.
• This method extends the applicability of high-level electronic structure calculations to medium-sized molecules and complex spectral simulations.