Anna Krylov of the University of Southern California is supported by an award from the Theory, Models and Computational Methods program to develop predictive ab initio methods to characterize the electronic structure of ionized and electron-attached states in model systems, such as small clusters of nucleobases, DNA hairpins or similar systems. The proposed developments are based on the equation-of-motion coupled-cluster approach for ionization potentials (EOM-IP) and electron-attached (EOM-EA) systems, which is capable of describing multiple closely lying electronic states in an accurate, robust and efficient computational scheme The PI and her research group are (i) implementing properties and analytic gradient calculations for EOM-IP-CCSD using a Frozen Natural Orbitals (FNO) approach); (ii) implementing non-iterative triples corrections for the IP-CCSD and IP-CISD methods to improve accuracy of these methods; (iii) extending the FNO scheme to EOM-EA-CCSD energies and gradient calculations; (iv) implementing efficient calculations of Dyson orbitals and other inter-state properties to aid interpretation of the results and facilitate more direct comparisons with the experiments; and (v) parallelizing and redesigning the tensor library to take advantage of modern computer architecture.

The Krylov group develops methods with the aim to study charge transport through molecular wires, which plays a central role in many important processes such as the radiative or oxidative damage in DNA, light harvesting, molecular electronics and solar energy applications. Software developed by this research group is made available to the community by being implemented in widely used software packages: Q-CHEM and SPARTAN.

This award is supported partially by a co-funding arrangement with the Office of Cyberinfrastructure (OCI).

Project Report

Charge transport (CT) through molecular wires plays the central role in a processes such as radiative or oxidative damage of DNA, light harvesting, molecular electronics and solar energy applications. The main project outcome is development of new predictive ab initio methods and computer codes (based on equation-of-motion coupled-cluster approach) that can be employed to characterize electronic structure of ionized and electron-attached states in model systems, such as small clusters of nucleobases, DNA hairpins, etc). These developments greatly increased the scope of problems that can be tackled computationally. We applied these tools to quantify the effects of pi-stacking, hydrogen-bonding, and micro-solvation on the IEs of nucleobases. The results of this work have been published in 22 peer-reviewed publication and one review article. Intellectual merit. Developing, benchmarking, and applying new many-electron models promotes our understanding of electron correlation and electronic structure. The latter is particularly rich in open-shell and electronically excited systems, where the complex interactions between unpaired electrons result in unusual bonding patterns. We developed predictive computational methods that are capable of describing open-shell species with accuracy comparable to that of state-of-the-art methodology available for closed-shell systems. By applying these tools to study prototypical CT systems, we learned more about fundamental aspects of CT. Quantitative understanding of the effects of non-covalent interactions on the IEs of individual nucleobases in small model systems (dimers, trimers, tetramers of nucleobases with and without water, etc) is an important prerequisite for theoretical studies of CT in more realistic systems. For example, these studies helped to characterize intrinsic properties of the building blocks of DNA separating them from the effects of the environment. Moreover, high-level calculations on these relatively small systems, which can be validated by the gas-phase experiments, serve as important benchmark sets for developing less computationally expensive methods, e.g., within density functional theory, polarizable force fields, or tight-binding models. Broader impact of our research includes training and mentoring of graduate students and postdocs for careers in academia and industry and contributions to the research infrastructure by integrating new computer codes in the widely used ab initio programs Q-CHEM and SPARTAN making them available to the broad chemistry community, as well as development of free open-source libraries (libtensor). This project provided training for students and postdocs in the field of theoretical and computational electronic structure. The students and postdocs learned how apply these fundamental concepts and tools to investigations of processes relevant to biology, materials, and energy research, thereby addressing problems of societal and technological importance. Our research also included a strong component of computer code design and implementation, preparing students for careers in high-tech industries where the ability to solve complex problems using sophisticated computational tools is highly valued. Participation in conferences, seminar programs and joint forums with other USC physical chemistry groups helped to develop broad vision as well as communication skills. The PI is also an active contributor to the Women in Science and Engineering (WISE) program at USC.

Agency
National Science Foundation (NSF)
Institute
Division of Chemistry (CHE)
Type
Standard Grant (Standard)
Application #
0951634
Program Officer
Evelyn M. Goldfield
Project Start
Project End
Budget Start
2010-09-01
Budget End
2013-08-31
Support Year
Fiscal Year
2009
Total Cost
$435,000
Indirect Cost
Name
University of Southern California
Department
Type
DUNS #
City
Los Angeles
State
CA
Country
United States
Zip Code
90089