Using Machine Learning to Understand the Causes of Quantum Decoherence in Solution-Phase Bond-Breaking Reactions

Area of Science:

• Quantum Chemistry
• Chemical Physics
• Computational Chemistry

Background:

• Decoherence is a fundamental quantum phenomenon limiting quantum computing.
• Environmental interactions cause quantum state collapse.
• Precise environmental chemical motions driving decoherence are not well understood.

Purpose of the Study:

• Investigate solvent-induced decoherence in photodissociation.
• Identify molecular mechanisms of decoherence during chemical bond breaking.
• Explore the role of solvent fluctuations in determining chemical bond breaking products.

Main Methods:

• Quantum molecular dynamics simulations of Na₂⁺ photodissociation in liquid Ar.
• Machine learning to analyze solute-solvent interactions in a high-dimensional feature space.
• Predicting electron localization onto photofragments based on environmental characteristics.

Main Results:

• Solvent fluctuations directly induce decoherence during Na₂⁺ photodissociation.
• Photofragment separation and out-of-phase solvent collisions are key factors in decoherence.
• Machine learning effectively characterizes complex solution-phase chemical processes.

Conclusions:

• Identified specific molecular motions causing decoherence in a chemical reaction.
• Demonstrated the utility of machine learning in interpreting complex chemical dynamics.
• Provided insights into the fundamental limits of quantum computing imposed by decoherence.