Wavefunction Optimization at the Complete Basis Set Limit with Multiwavelets and DMRG.

Area of Science:

• Quantum Chemistry
• Computational Physics
• Numerical Methods

Background:

• Strongly correlated quantum systems pose significant computational challenges.
• Traditional methods like Coupled Cluster and Configuration Interaction have limitations in accuracy and convergence.
• Multiwavelet-based multiresolution analysis (MRA) offers adaptive and hierarchical function representation.

Purpose of the Study:

• To develop an efficient algorithm combining DMRG and MRA for quantum systems.
• To leverage the strengths of both DMRG (multireference capability) and multiwavelets (complete basis set limit).
• To improve computational efficiency and accuracy in quantum chemistry calculations.

Main Methods:

• Integration of Density Matrix Renormalization Group (DMRG) within a multiwavelet-based multiresolution analysis (MRA).
• Adaptation of a Lagrangian optimization algorithm for MRA-represented orbitals.
• Replacement of Configuration Interaction (CI) calculations with DMRG for improved efficiency.
• Direct extraction of energy gradients from DMRG tensors, bypassing reduced density matrix computation.

Main Results:

• The combined DMRG-MRA approach successfully applied to small molecules (H2, He, HeH2, BeH2, N2).
• Achieved lower final energies compared to Full Configuration Interaction (FCI) on atomic orbital basis sets.
• Maintained a low number of orbitals while approaching the complete basis set limit.

Conclusions:

• The novel DMRG-MRA algorithm offers a computationally efficient and accurate method for strongly correlated quantum systems.
• This approach provides a powerful alternative to traditional quantum chemistry methods.
• The technique demonstrates potential for solving larger and more complex quantum problems.