Atom-Based Bootstrap Embedding For Molecules.

Area of Science:

• Quantum chemistry
• Computational chemistry
• Electronic structure theory

Background:

• Quantum embedding methods are crucial for accurate molecular simulations.
• Density Matrix Embedding Theory (DMET) faces limitations due to rigid system partitioning.
• Bootstrap Embedding (BE) offers flexibility with overlapping fragments but has orbital connectivity ambiguities.

Purpose of the Study:

• To introduce a novel atom-based fragment definition for Bootstrap Embedding (BE).
• To enhance the performance and accuracy of BE for molecular systems.
• To develop a low-scaling and highly accurate method for electron correlation in large molecules.

Main Methods:

• Developed an atom-based fragment definition for Bootstrap Embedding (BE).
• Applied the atom-based BE method to calculate valence electron correlation.
• Utilized extrapolation techniques to achieve near-chemical-accuracy.

Main Results:

• The atom-based BE method effectively recovers valence electron correlation.
• Achieved near-chemical-accuracy results in moderate-sized basis sets.
• Demonstrated significant augmentation of BE's performance in molecular applications.

Conclusions:

• Atom-based BE provides a more robust and less ambiguous approach to quantum embedding.
• The method shows promise for accurate and efficient electron correlation calculations in large molecules.
• Atom-based BE is anticipated to be a low-scaling and highly accurate computational chemistry approach.