CAREER: Structure Prediction and Design of Molecular Crystals with the GAtor Genetic Algorithm Package

点击翻译按钮获取中文摘要

NONTECHNICAL SUMMARYThe Division of Materials Research and the Division of Advanced Cyberinfrastructure contribute funds to this award. It supports computational research, algorithm development, and education on predicting properties of molecular crystals. Molecular crystals are used for many technological applications, including organic electronics, solar cells, and non-linear optics. They are of particular importance to the pharmaceutical industry because most drugs are marketed as molecular crystals of the pharmaceutically active ingredient. Molecular crystals are held together by van der Waals interactions between molecules. Unlike chemical bonds, van der Waals interactions arise from quantum mechanical fluctuations of the electron density that leads to a weak but long-ranged attractive electrostatic force. Because van der Waals interactions are weak, a given molecule may crystallize in more than one structure. Different crystal structures that could lead to molecular crystals that possess markedly different physical and chemical properties can all be nearly the most favored structure. Accurate calculations are required to predict the structure that is the winner of this competition and to be able to design new molecular crystals. Through the portal of computer simulations, the PI will gain access to the vast configuration space of materials structure and composition. She will explore the uncharted territory of molecular crystals that have not been synthesized yet and predict their properties knowing only their elemental composition and the laws of quantum mechanics. In this way, promising candidate molecular crystals for specific applications may be identified and used to guide synthesis efforts.Genetic algorithms (GAs) are guided to the most promising configurations by computer simulation that implements the "survival of the fittest" principle of evolution. Combining structural "genes" of the fittest structures in the population lead to "offspring" that propagate desirable features, while random mutations are employed to maintain diversity. This approach is implemented in a GA software package, GAtor, for computers that can perform many operations at the same time. The PI will distribute the GAtor software package to the broader community.This award also supports educational activities. The goals of the educational plan are to: (i) impart discipline-specific knowledge and technical skills; (ii) develop the foundational skills of science and engineering, such as problem solving, critical thinking, and computational thinking; (iii) instill professional skills, such as literature research, scientific writing, and presentation skills; (iv) prepare students for a career; and (v) broaden the participation of women and underrepresented minorities. This will be achieved via research mentoring, teaching of graduate and undergraduate courses, and the "Women Leaders in Physics and Engineering" seminar series.TECHNICAL SUMMARYThe Division of Materials Research and the Division of Advanced Cyberinfrastructure contribute funds to this award. It supports computational research, algorithm development, and education on molecular crystals. The goal of the proposed research is computational structure prediction and design of molecular crystals with tailored properties from first principles. This will be achieved by developing a versatile genetic algorithm (GA) package, GAtor, capable of conducting both energy-based and property-based optimization with high accuracy and high efficiency. The research plan comprises two phases: (I) the Structure Prediction phase seeks to predict the most stable structure(s) of a molecular crystal, given a "stick diagram" of a single molecule- a grand challenge embodied by the Crystal Structure Prediction blind test; and (II) the Crystal Engineering from First Principles phase seeks to design molecular crystals with target properties for applications in organic electronics and photovoltaics by property-based GAs- an emerging paradigm in computational materials design. The PI will use density functional theory based simulations that include dispersion to accurately describe the potential energy landscape arising from van der Waals interactions. Many-body perturbation theory methods will be used to accurately describe excitonic properties. The educational component is aimed to: (i) impart discipline-specific knowledge and technical skills; (ii) develop the foundational skills of science and engineering, such as problem solving, critical thinking, and computational thinking; (iii) instill professional skills, such as literature research, scientific writing, and presentation skills; (iv) prepare students for a career; and (v) broaden the participation of women and underrepresented minorities. This will be achieved via research mentoring, teaching of graduate and undergraduate courses, and the "Women Leaders in Physics and Engineering" seminar series.