Non-Empirical and Self-Interaction Corrections for DFTB: Towards Accurate Quantum Simulations for Large Mesoscale Systems (Final Report)
This project was comprised of two complementary (but parallel) thrusts: (1) implementing massively-parallelized computing hardware (with new computational hardware that may replace GPUs) in the density functional tight binding (DFTB) approach and (2) developing new capabilities in DFTB to calculate the electronic structure and dynamics of large chemical systems. While classical molecular dynamics can handle hundreds of thousands of atoms, it cannot provide a first-principles based description of chemical systems at the quantum level. At the other extreme, conventional Kohn-Sham DFT methods can probe the true quantum mechanical nature of chemical systems; however, these methods cannot tackle the large sizes and length scales relevant to dynamics simulations of realistic systems. The DFTB formalism utilized in this project provides a viable approach for probing these large systems at a quantum mechanical level of detail. However, to utilize the DFTB approach for accurate calculations of electronic properties, it is crucial to incorporate quantum-based corrections in DFTB since exchange-correlation effects can still remain very strong in these large systems. At the same time, enhancing the computational efficiency of DFTB is also essential since optimal computational performance is required for addressing the large size scales associated with realistic chemical systems. As such, the new non-empirical corrections and computing hardware enhancements implemented in this project will enable accurate and computationally efficient approaches to directly probe electronic properties in these large, complex systems.
- Research Organization:
- Univ. of California, Riverside, CA (United States)
- Sponsoring Organization:
- USDOE Office of Science (SC), Basic Energy Sciences (BES). Chemical Sciences, Geosciences & Biosciences Division (CSGB)
- DOE Contract Number:
- SC0016269
- OSTI ID:
- 1842676
- Report Number(s):
- DOE-UCR-16269; TRN: US2403509
- Country of Publication:
- United States
- Language:
- English
Similar Records
Heterogeneous CPU+GPU-Enabled Simulations for DFTB Molecular Dynamics of Large Chemical and Biological Systems
ORNL_AISD-Ex: Quantum chemical prediction of UV/Vis absorption spectra for over 10 million organic molecules