Incorporating Linear Synchronous Transit Interpolation into the Growing String Method: Algorithm and Applications.

Area of Science:

• Theoretical Chemistry
• Computational Chemistry
• Chemical Dynamics

Background:

• The growing string method aids in identifying transition states for chemical reactions.
• Current interpolation schemes, like cubic splines in Cartesian coordinates, can be inefficient, requiring numerous optimization steps.
• This inefficiency leads to increased computational cost due to repeated energy and gradient calculations.

Purpose of the Study:

• To develop a more efficient interpolation and reparameterization method for the growing string technique.
• To reduce the computational cost associated with identifying reaction pathways and transition states.
• To improve the accuracy and speed of theoretical chemical reaction studies.

Main Methods:

• Implementation of the linear synchronous transit (LST) method for node interpolation and reparameterization.
• Integration of the LST method within the growing string framework.
• Application to benchmark chemical systems, including the alanine dipeptide rearrangement and a cationic alkyl ring condensation.

Main Results:

• Achieved a significant speedup in computational cost, ranging from 30% to 50%.
• Demonstrated the effectiveness of the LST-based interpolation for generating more accurate intermediate structures.
• Reduced the number of optimization steps required to converge to the reaction path.

Conclusions:

• The proposed LST-based interpolation method enhances the efficiency of the growing string method.
• This advancement offers a computationally cheaper alternative for theoretical studies of chemical reactions.
• The method shows promise for broader application in computational chemistry for transition state searches.