Daniel Gibney

Chemistry Postdoctoral Scholar · University of Minnesota

I am currently a Postdoctoral Scholar with Professor Boyn at the University of Minnesota working on both the development and application of physics-based reduced density matrix approaches to predicting chemical properties. My research areas involve:

1. The intersection of Density Functional Theory (DFT) and 1-Electron Reduced Density Matrix Functional Theory (1-RDMFT) to capture strong correlation effects while retaining DFT's favorable computational scaling. I have developed a method termed hybrid DFA 1-RDMFT that allows for strong correlation effects to be captured through a 1-RDM functional while the remaining weak correlation is described through standard density functionals. This approach retains the computational scaling of the underlying density functional and produces orbital occupations in line with CASSCF calculations in strongly correlated systems. In weakly correlated systems, it is equivilent to DFT.
2. The 2-electron reduced density matrix (2-RDM) based Anti-hermitian Contracted Schrodinger Equation (ACSE). This method shows great promise in being capable of solving the Schrodinger equation within milihartree accuracy across a wide domain of varied chemical problems for both ground and excited states. Furthermore, it is capable of being started from a strongly correlated CASSCF reference wavefunction to recover all-electron correlation. Here, the ACSE enables the recovery of all-electron correlation in systems with many strongly correlated electrons, which is generally intractable with often used MPRT2 methods due to their high computational scaling with the size of the active space. As the ACSE operates on the all electron 2-RDM, it has no computational dependence on the size of the active space.

   Recent work on this front has involved the creation of an open source python based implementation to act as a platform for future development. Ongoing work broadly involves the acceleration of the ACSE through increased computational efficiency and potential methods for reducing computational scaling, application of alternative reconstruction techniques, and methods for reducing propagation error.

3. The application of high throughput DFT calculations