Machine Learning for Electronically Excited States of Molecules.

Area of Science:

• Computational Chemistry
• Quantum Mechanics
• Artificial Intelligence

Background:

• Electronically excited states of molecules are crucial for photochemistry, photophysics, photobiology, and material science.
• Accurate theoretical descriptions of these states rely on computationally intensive quantum chemical calculations.

Purpose of the Study:

• To review the application of machine learning (ML) in accelerating excited-state simulations.
• To explore how AI can advance research in excited-state molecular science.
• To discuss the promises and challenges of ML techniques in this field.

Main Methods:

• Review of existing literature on ML applications for excited states.
• Discussion of quantum chemistry methods for electronic structure and non-adiabatic dynamics.
• Analysis of ML techniques used in conjunction with quantum chemistry calculations.

Main Results:

• ML significantly speeds up simulations of excited-state dynamics and static properties like absorption spectra.
• Various ML approaches are being integrated with quantum chemistry for excited-state studies.
• The review contextualizes ML applications by detailing underlying computational methods and their limitations.

Conclusions:

• Machine learning offers powerful tools to overcome the computational cost of studying excited molecular states.
• Understanding the interplay between ML and quantum chemistry is key to future advancements.
• Careful consideration of ML technique promises and pitfalls is essential for reliable results.