Preface: Special Topic on Reaction Pathways Free

In the last few decades, our ability to generate high resolution data by computer simulation has dramatically increased. In the field of biophysics, it is now possible to use molecular dynamics (MD) simulation to fold small proteins,^1,2 study large conformational changes in macromolecules,³ or protein-ligand and protein-protein association⁴ in silico. In the fields of chemistry and material science, where bond-breaking reactions are of particular interest, there have been significant advances in the accuracy and scale of electronic structure calculations of activated processes, with applications to understanding materials such as batteries⁵ and catalysts.⁶ 

The advances in simulation have in turn stimulated a surge of interest in data analysis methods,⁷ to extract essential information from the data in order to enable quantitative predictions and connect with experimental results. In particular, both in biophysics and in material science, it is desirable to characterize complex molecular mechanisms in terms of reaction pathways.⁸ The systems of interest are generally composed by a large number of atoms, and the analysis of the reaction mechanisms requires a reduction in the dimensionality of the simulation data to a few collective variables that describe the molecular processes under consideration.

In this Special Topic Issue of The Journal of Chemical Physics, we have compiled examples of work where the definition of collective variables and reaction pathways is essential to understand molecular mechanisms, with application both in biophysics and in material science.

While methods developed in these two different sub-fields of chemistry bear significant similarities and have sometimes a common origin, there are also important differences associated with the different application domains. One important distinction among these methods is based on the requirement of the knowledge of (meta)stable states in the system of interest. In the field of biophysics, it is desirable to sample the phase space relevant for a molecular system, to map all the main metastable states and the rates of interconversion between them. For a material scientist, it is very often the time evolution of a system which is important, including the final states which are not known a priori. In catalysis, for example, one typically knows the initial structure of the catalysts and the reactions, and what you want to know is the active site under steady state conditions and the products that form.

In both biophysical and material science systems, the definition of reaction coordinates and reaction pathways typically rely on the existence of a separation of time scales.⁹ The slow processes are usually of interest, and most methods propose to find the important slow dynamical “directions” while averaging over the faster processes. Transition state theory, for example, can be used to separate the fast vibrational time scale from the slower state-to-state kinetics. For the rough landscapes found in biophysical systems, the use of collective variables, spectral methods, Markov state models, and transition path theory is central to a description of the slowest time evolution.^7,10,11 Here, we recognize that there are similarities as systems in both fields become increasingly complex, and an aim of this Special Issue is to explore common ground between methods used in what could appear to be very diverse fields.

C.C. group is supported by NSF (Grant Nos. CHE-1265929, CHE-1738990, and PHY-1427654) and the Welch Foundation (Grant No. C-1570). G.H. group is supported by NSF (Grant Nos. CHE-1534177, CHE-1505135, and DMR-1410335) and the Welch Foundation (Grant No. F-1841).