Dynamical pruning of the multiconfiguration time-dependent Hartree (DP-MCTDH) method: An efficient approach for...

Area of Science:

• Quantum dynamics
• Computational chemistry
• Theoretical physics

Background:

• The multiconfiguration time-dependent Hartree (MCTDH) method is a powerful tool for simulating quantum dynamics.
• Dynamical pruning (DP) aims to improve computational efficiency by selecting relevant basis functions on-the-fly.
• Combining DP with MCTDH (DP-MCTDH) offers potential for handling larger and more complex quantum systems.

Purpose of the Study:

• To present and evaluate two novel strategies for integrating dynamical pruning with the MCTDH method.
• To demonstrate the applicability and benefits of these DP-MCTDH strategies for molecular systems.
• To assess the computational speed-ups and performance improvements offered by the proposed methods.

Main Methods:

• The study introduces two distinct DP-MCTDH strategies.
• Strategy 1: Pruning the primitive basis set representing single-particle functions (SPFs).
• Strategy 2: Pruning the set of SPF configurations at each time step.

Main Results:

• The first strategy is effective for smaller systems (e.g., NO2) requiring numerous basis functions per degree of freedom.
• This strategy enables higher-dimensional mode combinations and relaxes the strict sum-of-product form requirement for Hamiltonians, as shown for 24-dimensional pyrazine.
• The second strategy achieves significant computational speed-ups, ranging from 5 to 50 times, making it competitive with the multilayer MCTDH approach.

Conclusions:

• The presented DP-MCTDH strategies offer efficient computational approaches for quantum dynamics simulations.
• These methods provide significant advantages in terms of speed and applicability to higher-dimensional systems.
• The findings suggest that DP-MCTDH is a promising alternative to existing methods like multilayer MCTDH for complex quantum mechanical problems.