Event details from: Massively parallel path sampling for ab initio aqueous chemical reactions

Date
12 June 2025
Time
14:00

Venue

Room A1.28

Speaker

Prof.dr. Titus van Erp (NTNU, Trondheim)

Abstract

Path sampling is a combination of Monte Carlo (MC) and molecular dynamics (MD) that serves as an enhanced sampling technique for studying molecular transitions. These transitions cannot be efficiently studied using straightforward MD simulations due to the computational time required, which could amount to centuries. Unlike most other enhanced sampling methods, path sampling enables the exact study of dynamics without modifying them or the potential energy surface.
At the core of path sampling is the shooting move [1], an MC move that generates a new path from a previous one. Through replica exchange transition interface sampling (RETIS) [2], multiple path sampling simulations can be combined, each focusing on specific path ensembles that reflect different stages of the reaction, allowing the computation of rates and other mechanistic outputs.
Although RETIS is highly efficient in terms of CPU usage, it is difficult to parallelize because paths have varying lengths and must be completed before they can participate in replica exchange. Recently, we developed the Infinity-RETIS approach [3], which employs asynchronous replica exchange and infinite swaps, complemented by advanced shooting moves that decorrelate rapidly and have high acceptance rates. Our recent work demonstrates that this method enables massively parallel path sampling, yielding results within days instead of months.
In this talk, I will discuss two recent ab initio applications of Infinity-RETIS: redox reactions involving electron transfer between positively charged ruthenium ions and the formation of carbonic acid, both in aqueous solution.

[1] C. Dellago, P. G. Bolhuis, F. S. Csajka, D. Chandler, Transition path sampling and the calculation of rate constants.J. Chem. Phys. 108, 1964 (1998).
[2] T. van Erp, Reaction rate calculation by parallel path swapping. Phys. Rev. Lett. 98, 268301 (2007)
[3] D. T. Zhang, L. B. S. Roet, A. Lervik, and T. S. van Erp, Highly parallelizable path sampling with minimal rejections using asynchronous replica exchange and infinite swaps. Proc. Natl. Acad. Sci. U.S.A. 121, e2318731121 (2024).

Contact