Jankunas Dissertation Award (Finalists): Practical Electronic Structure Methods for Strongly Correlated Electron Simulations

Understanding quantum many-body effects arising from strong electron correlations has been a long-standing challenge in chemical physics. In particular, making quantitative, material-specific predictions of electronic and magnetic properties of large, realistic systems is key to designing next-generation high-temperature superconductors for energy applications and molecular spin qubits for quantum information science.

During my Ph.D., I systematically studied Kondo physics, a prototypical many-body quantum phenomenon, in transition metal impurities. Since its experimental characterization six decades ago, it has remained difficult for theory to qualitatively capture the element-specific trend. Utilizing our recently developed ab initio dynamical mean-field theory, we were able to converge to an exact zero-temperature electronic treatment of Kondo correlations. Our converged results captured the subtle exponential trend of the Kondo temperature in experiment across a series of 3d transition metals. This demonstrates the predictive power for ab initio quantum many-body simulation and more broadly the potential towards a systematically improvable simulation of experimental observables of correlated electron phenomena.

In a separate effort to design lanthanide single-ion magnets, I theoretically predict effective crystal field Hamiltonians that provide the key description of the electronic and magnetic properties. We introduced a simple approach to derive Hamiltonian parameters using density functional calculations of randomly rotated mean-field states within the low-energy manifold. Benchmarked on five lanthanide single-ion magnets, our method consistently achieves accuracy comparable to the state-of-the-art from correlated wavefunction methods, but with a significantly reduced computational cost. Its low mean-field scaling has enabled predictions of systems with up to 186 atoms, more than doubling the previous size limit. We have further applied this method to designing a new clock qubit array in a metal-organic framework, where our theory captures all salient aspects of the magnetic and EPR measurement.