Incorporation of Internal Coordinates Interpolation into the Freezing String Method.

Area of Science:

• Computational Chemistry
• Chemical Reaction Dynamics

Background:

• Accurate determination of transition states is crucial for understanding chemical reaction mechanisms.
• Existing methods for generating initial guesses for transition state searches can be computationally expensive and may fail for complex systems.

Purpose of the Study:

• To develop and validate an improved freezing string method (FSM) for more reliable and efficient transition state (TS) searches.
• To reduce the computational cost associated with transition state calculations.

Main Methods:

• Incorporation of internal coordinate interpolation into the freezing string method (FSM).
• Testing the enhanced FSM on over 40 diverse chemical reactions across three benchmark datasets.
• Comparison with methods using linear synchronous transit (LST) interpolation.

Main Results:

• The enhanced FSM demonstrates improved reliability and enables larger interpolation step sizes.
• A reduction in computational cost by nearly 50% was achieved due to fewer optimization steps per cycle.
• A 100% success rate was maintained on benchmark chemical reaction test cases, including challenging systems where LST failed.

Conclusions:

• The internal coordinate interpolation significantly enhances the performance and efficiency of the FSM for transition state searches.
• The developed method offers a robust and computationally cheaper alternative for exploring reaction pathways.
• An open-source Python implementation of the FSM and associated data are provided for community use.