AIQM2: organic reaction simulations beyond DFT.

Area of Science:

• Computational chemistry
• Quantum mechanics
• Artificial intelligence in science

Background:

• Density functional theory (DFT) is widely used for reaction simulations but faces limitations in cost and accuracy for large systems.
• Achieving high accuracy and computational efficiency simultaneously remains a challenge in computational chemistry.

Purpose of the Study:

• To introduce AIQM2, a universal AI-enhanced quantum mechanics method for fast and accurate large-scale organic reaction simulations.
• To demonstrate AIQM2's capability to overcome the limitations of traditional DFT methods for complex chemical reactions.

Main Methods:

• Development and application of AIQM2, an AI-enhanced quantum mechanics approach.
• Comparison of AIQM2 performance against standard DFT methods in terms of speed and accuracy.
• Utilizing AIQM2 for extensive reaction dynamics studies and mechanism elucidation.

Main Results:

• AIQM2 achieves speeds orders of magnitude faster than common DFT methods.
• The accuracy of AIQM2 in reaction energies, transition states, and barrier heights is comparable to DFT and approaches coupled cluster accuracy.
• AIQM2 demonstrates high transferability and robustness, outperforming pure machine learning potentials.

Conclusions:

• AIQM2 represents a breakthrough in enabling fast, accurate, and large-scale organic reaction simulations.
• The method allows for simulations at system sizes and time scales beyond current DFT capabilities.
• AIQM2 facilitates the revision of reaction mechanisms and product distributions through efficient computational studies.