Second-Order Active-Space Embedding Theory.

Area of Science:

• Quantum chemistry
• Computational chemistry
• Theoretical chemistry

Background:

• Quantum embedding schemes offer a path to extend multireference computations.
• Strong correlation effects in large molecules are often localized.
• Accurate treatment of these effects is computationally demanding.

Purpose of the Study:

• Introduce a new second-order active-space embedding theory (ASET(2)).
• Improve upon existing mean-field frozen embedding methods.
• Enhance the accuracy of quantum computations for strongly correlated systems.

Main Methods:

• Developed ASET(2) using an approximate canonical transformation.
• Formulated the canonical transformation via the driven similarity renormalization group.
• Benchmarked ASET(2) on N═N bond dissociation, S0 to S1 excitation, and complex interaction energies.

Main Results:

• ASET(2) generally improves the accuracy of embedded computations.
• Explicit treatment of fragment-environment interactions beyond mean-field is beneficial.
• ASET(2) is necessary for accurate excitation energies and singlet-triplet gaps in tested systems.

Conclusions:

• ASET(2) provides a more accurate description of localized strong correlation effects.
• The method advances the capability of quantum embedding for complex molecular systems.
• This approach is vital for accurate predictions in challenging chemical problems.