High-throughput ab initio reaction mechanism exploration in the cloud with automated multi-reference validation.

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Materials Science

Background:

• Quantum chemical calculations are essential for molecular matter studies but require significant manual input and expertise.
• Automation of these calculations can lower the barrier to entry for researchers and accelerate discovery.

Purpose of the Study:

• To present AutoRXN, an automated workflow for high-throughput electronic structure calculations.
• To enable autonomous reaction mechanism exploration with minimal operator interference.

Main Methods:

• Utilizes density functional theory (DFT) for initial structure optimization and energy calculations.
• Employs coupled cluster (CC) methods for accurate energy and property estimation of optimized structures.
• Integrates multi-reference diagnostics and automated multi-configurational calculations to validate CC results.

Main Results:

• Demonstrates the capability of AutoRXN for autonomous reaction mechanism exploration.
• Successfully applied to study the mode of action of a homogeneous catalyst for asymmetric ketone reduction.
• Highlights the workflow's autonomy, stability, and efficiency in cloud-based computational campaigns.

Conclusions:

• AutoRXN significantly streamlines and automates advanced quantum chemical calculations.
• The workflow democratizes access to high-throughput computational chemistry, reducing reliance on specialized expertise.
• AutoRXN facilitates large-scale computational studies, accelerating scientific discovery in molecular systems.