OMNI-P2x universal neural network potential for excited-state simulations

Area of Science:

• Computational Chemistry
• Materials Science
• Quantum Mechanics

Background:

• Photo-active molecular systems are crucial for technologies like solar cells and OLEDs.
• Understanding photophysical processes is key for designing new photo-responsive molecules.
• First-principles quantum-mechanical calculations are accurate but computationally expensive for high-throughput studies.

Purpose of the Study:

• Introduce OMNI-P2x, a universal neural network potential for molecular excited and ground electronic states.
• Provide a computationally efficient alternative to traditional methods for photophysical and photochemical simulations.
• Enable faster and more accurate rational design of photo-responsive materials.

Main Methods:

• Developed OMNI-P2x, a novel neural network potential.
• Utilized OMNI-P2x for UV/Vis absorption spectroscopy simulations.
• Performed real-time photodynamical simulations.
• Applied OMNI-P2x to the design of visible-light-absorbing azobenzene systems.

Main Results:

• OMNI-P2x achieves accuracy comparable to time-dependent density functional theory (TD-DFT) methods.
• OMNI-P2x offers a significant reduction in computational cost compared to TD-DFT.
• OMNI-P2x outperforms established semiempirical methods in speed and accuracy for excited-state simulations.
• Demonstrated successful application in spectroscopy, dynamics, and molecular design.

Conclusions:

• OMNI-P2x is a powerful and efficient tool for a wide range of photophysical and photochemical simulations.
• The neural network potential accelerates the discovery and design of advanced photo-active materials.
• OMNI-P2x bridges the gap between accuracy and computational cost in excited-state calculations.