Direct quantum dynamics using variational Gaussian wavepackets and Gaussian process regression

Area of Science:

• Quantum chemistry
• Computational physics
• Molecular dynamics

Background:

• Accurate simulation of quantum nuclear dynamics in anharmonic molecular systems is computationally demanding.
• Existing methods often struggle with efficiency and scalability for multidimensional systems.

Purpose of the Study:

• To present a novel method for direct variational quantum nuclear dynamics.
• To enable efficient and exact quantum dynamics simulations of complex molecular systems.
• To demonstrate the method's applicability to intramolecular proton transfer.

Main Methods:

• Direct variational quantum nuclear dynamics using Gaussian wavepackets.
• On-the-fly potential energy surface fitting with Gaussian process regression.
• Analytic evaluation of Hamiltonian matrix elements.

Main Results:

• The implemented method allows for black-box quantum dynamics simulations.
• Demonstrated successful application to intramolecular proton transfer in salicylaldimine.
• The approach provides exact and efficient calculations for multidimensional anharmonic systems.

Conclusions:

• The developed method offers a powerful tool for studying quantum nuclear dynamics.
• Future work will focus on algorithmic improvements and non-adiabatic dynamics.
• This approach has significant potential for simulating complex molecular processes.