Postdoc Work

(Prof. Sharon Hammes-Schiffer Group)

Quantum proton in FHF molecule

My current research project is mainly on the method development in the multicomponent nuclear-electronic orbital (NEO) approach. This approach treats both electrons and some key nuclei quantum-mechanically and describes them with the molecular orbital theory. In most cases of interest, specific key hydrogen nuclei will be treated quantum-mechanically when their quantum effects are important in a chemical process.

On-axis and off-axis proton densities for FHF molecule

Using my knowledge from my Ph.D. study, I developed an electron-proton correlation (epc) functional within the framework of NEO density functional theory (NEO-DFT). It is the first working epc functional, and it can accurately reproduce the delocalized proton density. We name the functional as epc17-1. I also reparametrized the functional and make it give good proton affinities, and this new form is named as epc17-2. Currently a better functional that can accurately describe both densities and energies is under development.

Proton orbitals for ground and vibrational excited states of FHF molecule

Furthermore, I derived the time dependent version of NEO-DFT, which is named as NEO-TDDFT. It can provide good proton vibrational energies as well as electronic excitations.

Ph.D. Work

(Prof. Weitao Yang Group)

My research project in Ph.D. was mainly on the particle-particle random phase approximation (pp-RPA). The pp-RPA has been a textbook method for treating correlation in nuclear physics for a long time. Recently, we introduced it to atomic and molecular systems. We also successfully combined it with density functional approximations besides the traditional Hartree-Fock approximation.

Heat of formation obtained from pp-RPA on G2/97 test set

The pp-RPA calculates excitation energies by calculating the difference of two-electron addition energies

We also applied the pp-RPA to electronic excitation problems. These problems include challenging double excitations, charge transfer excitations, Rydberg excitations, diradical problems, and conical intersections. The calculation usually starts from a two-electron deficient system and then recovers a series of neutral states by adding two electrons back to the system. Although it intrinsically misses those excitations originated from below the highest occupied molecular orbital, it can solve challenging excitation problems as mentioned above, which cannot be well described by the widely-used adiabatic time-dependent density functional theory (TDDFT).