Research

Computational Chemistry: ​Applications, Theory, and Software​

Main Research Interests

• Enzyme Reactions. There are many mysteries surrounding enzymes, such as CRISPR Cas9/12 proteins (what are the mechanisms for the target-strand and non-target-strand cleavages? How do mutations affect their catalytic efficiency and off-target cleavage rate?); Kemp eliminases (how did the laboratory-directed evolutions help improve the efficiency of these synthetic enzymes?); and DNA polymerase eta (what is the role of the third Mg2+ ion?). To solve these puzzles conclusively, we need to construct and analyze the reaction free energy landscape for each of these enzymes. But a free energy simulation of a given enzyme requires the sampling of millions of configurations, and an accurate description (using density-functional theory or better models) of the active pocket of each configuration would be quite costly (>500,000 CPU hours for the entire simulation!). To reduce this bottleneck, we recently made two methodological advances: (1) the first machine-learning potential for capturing the energetics of enzyme systems [1]; and (2) an improved multiple timestep simulation protocol [2]. These new tools greatly facilitate our ongoing modeling of enzyme systems, especially CRISPR Cas proteins (with Rakhi Rajan and Jin Liu) and IPpoI (with Yang Gao). ​

• Chemical Reactions. Of particular interest to us are oxygen evolution and chemiluminescence reactions, where the reactant and product have different spin-multiplicities (such as singlet->triplet). For these reactions, we can identify the minimum energy crossing points (MECP) using the algorithm from Harvey and others. But an important issue remains unclear, namely how probable it is for a system to hop between different spin surfaces at the MECP, which is a key factor of the overall reaction rate. Our strategy is to continue our development of the spin-adiabatic approach for spin-crossing reaction studies [3]. It will allow us to effectively model ion-molecule reactions (with Zhibo Yang and Peter Armtrout), N2O splitting (with Ken Nicholas), chemiluminescence probes for Alzheimer’s disease research (with Chongzhao Ran), and firefly bioluminescence reaction efficiency.

• Polariton Chemistry. Polaritons -- hybrid light-matter states -- form in Fabry-Perot photon cavities. We are especially interested in exciton polaritons, which couple a collective electronic excitation over a larger number (N) of molecules to a resonant (or near-resonant) cavity mode. It is fascinating that one can tune the energy of plaritonic states by varying the number density (N/V), which offers a new venue to modulate the rate of photoisomerization and other photochemical/physical processes. We have contributed to the field by implementing the cavity quantum electrodynamics time-dependent density functional theory methods [4,5] and proposing an alternative mechanism for effective reverse intersystem crossing [6]. Now we are working towards addressing a super-challenging question, namely for experimental systems of interest, what is the role of the N​-1 dark states, especially the rate of transition into and out of these dark states?