Geodesic interpolation for reaction pathways.

Area of Science:

• Computational chemistry
• Chemical reaction dynamics
• Reaction path optimization

Background:

• High-throughput reaction discovery relies on accurate minimum energy path calculations.
• Standard interpolation methods in Cartesian coordinates often yield poor approximations and slow convergence.
• Efficient reaction path generation is crucial for computational chemistry and materials science.

Purpose of the Study:

• To develop a more robust and efficient method for interpolating between reactant and product geometries.
• To improve the starting guesses for minimum energy path calculations in computational chemistry.
• To address the limitations of traditional interpolation techniques in reaction discovery.

Main Methods:

• Reformulating geometry interpolation as a geodesic curve search on a Riemannian manifold.
• Explicitly accounting for the metric tensor to correct for coordinate redundancies.
• Utilizing geometric information to generate initial reaction paths.

Main Results:

• Demonstrated that coordinate choices implicitly alter the metric, affecting interpolation performance.
• Developed a method that achieves accurate interpolation in Cartesian coordinates by explicitly managing the metric.
• Generated geodesic paths that closely approximate true minimum energy paths.

Conclusions:

• The geodesic path approach provides excellent starting points for minimum energy path algorithms.
• This method enhances the efficiency of reaction discovery and rate calculations.
• The findings offer a significant improvement for computational chemistry workflows.