Communication: Active space decomposition with multiple sites: density matrix renormalization group algorithm.

Area of Science:

• Quantum chemistry
• Computational physics
• Materials science

Background:

• The active space decomposition method is a computational technique for simplifying electronic structure calculations.
• Accurate modeling of electron correlation is crucial for understanding molecular properties.
• Extending existing methods to handle more complex systems is an ongoing challenge.

Purpose of the Study:

• To generalize the active space decomposition method for systems with more than two active sites.
• To investigate the efficiency and accuracy of the extended method using the density matrix renormalization group algorithm.
• To assess the method's performance on realistic molecular models.

Main Methods:

• Extension of the active space decomposition method to multiple active sites.
• Utilizing the density matrix renormalization group (DMRG) algorithm for electron correlation.
• Employing complete or restricted active-space wave functions for fragment descriptions.
• Numerical calculations on benzene pentamer and perylene diimide trimer models.

Main Results:

• The extended method demonstrates a near-exponential decrease in truncation errors with an increasing number of renormalization states (M).
• Numerically exact calculations (within a few μE(h)) were achieved with M = 128 for both benzene pentamer and perylene diimide trimer.
• The rapid convergence is attributed to the targeted application of renormalization steps to interfragment electron correlation.

Conclusions:

• The generalized active space decomposition method is highly effective for accurate electronic structure calculations of complex molecules.
• The density matrix renormalization group algorithm provides efficient and accurate treatment of interfragment electron correlation.
• This method offers a promising pathway for computationally exact studies of large molecular systems.