Tensor network simulation of multi-environmental open quantum dynamics via machine learning and entanglement...

Area of Science:

• Quantum dynamics simulation
• Computational chemistry
• Molecular physics

Background:

• Simulating open quantum dynamics is crucial for developing new quantum devices.
• Strong, non-Markovian coupling between molecular vibrations and electronic states presents a significant challenge.
• Existing methods struggle with the complexity of large vibronic wave functions.

Purpose of the Study:

• To present a general computational strategy for simulating complex molecular quantum dynamics.
• To enable accurate real-time computation of non-perturbative vibronic dynamics.
• To provide direct visualization of dissipative photophysics in molecular systems.

Main Methods:

• Utilized tensor network methods for efficient computation of large vibronic wave functions.
• Integrated ab initio modeling and machine learning for enhanced simulation capabilities.
• Employed entanglement analysis to understand quantum system-environment interactions.

Main Results:

• Developed an advanced computational strategy for simulating non-perturbative, real-time quantum dynamics.
• Successfully computed dynamics of exponentially large vibronic wave functions for real molecules.
• Enabled real-time insight and direct visualization of dissipative photophysics.

Conclusions:

• The presented strategy effectively overcomes challenges in simulating open quantum dynamics.
• This approach provides unprecedented real-time insight into molecular quantum processes.
• Demonstrated applicability with an example of singlet fission, paving the way for future device design.