Abstract:
Accurate quantum chemistry prediction of electron correlation energies is prohibitively expensive for complex molecules when explicitly solving post-Hartree-Fock equations. In several limits of approximation, efficient low-scaling correlated electronic structure methods have been established for molecules and solids [1, 2]. These methods have been demonstrated to often achieve remarkable accuracy with significantly reduced computational costs by orders of magnitude. I will discuss our recent developments [3-10] that push and leverage the limit of scalable correlated electronic structure theories, including 1) the CPU/GPU-enabled parallel scalable methods; 2) the analytical gradient theory for driving ab initio Born-Oppenheimer molecular dynamics; 3) the quantum-feature-informed deep neural network model for making transferable prediction across varying molecular and crystalline scales; 4) the quantum-polarizable ab initio neural network water model for simulating liquid water and ice systems; as well as 5) an effective configuration interaction Hamiltonian for treating strongly correlated states.
Biography:
Dr. Jun Yang is a tenured associate professor of theoretical chemistry in the department of chemistry at the University of Hong Kong. He had BSc and MSc both at Peking University, PhD at University of Cologne with Professor Michael Dolg, and postdoced at Cornell/Princeton Universities with Professor Garnet Chan, before joining HKU as tenure-track assistant professor in 2016. He was awarded DFG fellowship to start his PhD in 2004, the IBM-Löwdin Postdoc Award in 2010 and Hong Kong RGC Early Career Award in 2017. Dr. Yang’s main research interests remain on scalable, precise and intelligent electronic structure methodologies and applications. His work has been published and featured in various core journals, including JCTC, JCP, Science, PNAS, JACS and ANIE.
Seminar Series by the NYU-ECNU Center for Computational Chemistry at NYU Shanghai
This event is open to the NYU Shanghai, NYU, ECNU community and the computational chemistry community.