These are preliminary reports that have not been peer-reviewed. They should not be regarded as conclusive, guide clinical practice/health-related behavior, or be reported in news media as established information. For more information, please see our FAQs.
2 files

Accurate, Automated Density Functional Theory for Complex Molecules Using On-the-fly Error Correction

submitted on 09.10.2020, 20:56 and posted on 12.10.2020, 12:56 by Samuel Blau, Evan Walter Clark Spotte-Smith, Brandon Wood, Shyam Dwaraknath, Kristin Persson
High-throughput density functional theory (DFT) has been widely utilized to study a variety of materials and molecular properties. However, its application to complex molecular systems, including those relevant to electrochemical reactivity and decomposition, has been limited by insufficient automation.Here, we report a broadly applicable, automated framework for the accurate and robust DFT calculation of molecules, capable of addressing species relevant to electrochemistry. This framework is specifically designed to study molecules with different charge states, open-shell electronic structure, metal coordination, and implicit solvation. We first identify appropriate levels of theory that avoid calculation failures and accurately predict molecular redox potentials. We then describe our framework, including methods to automatically detect and correct errors and to optimize structures from saddle points to potential energy surface minima. To demonstrate the efficacy of this framework, we examine a case study including over 12,000 calculations of reactive molecular fragments. This framework is able to reduce the rate of failure for DFT calculations from 25.1% to 1.2%, significantly improving the degree of automation possible for high-throughput molecular DFT.


Email Address of Submitting Author


Lawrence Berkeley National Laboratory



ORCID For Submitting Author


Declaration of Conflict of Interest

The authors have no conflicts of interest to declare.