UC Berkeley

Organic molecules are versatile and tunable building blocks for technology, in nanoscale and bulk devices. In this dissertation, I will consider some important applications for organic molecules involving optical and transport properties, and develop methods and software appropriate for theoretical calculations of these properties. Specifically, we will consider second-harmonic generation, a nonlinear optical process; photoisomerization, in which absorption of light leads to mechanical motion; charge transport in junctions formed of single molecules; and optical excitations in pentacene, an organic semiconductor with applications in photovoltaics, optoelectronics, and flexible electronics. In the Introduction (Chapter 1), I will give an overview of some phenomenology about organic molecules and these application areas, and discuss the basics of the theoretical methodology I will use: density-functional theory (DFT), time-dependent density-functional theory (TDDFT), and many-body perturbation theory based on the GW approximation. In the subsequent chapters, I will further discuss, develop, and apply this methodology.

2. I will give a pedagogical derivation of the methods for calculating response properties in TDDFT, with particular focus on the Sternheimer equation, as will be used in subsequent chapters. I will review the many different response properties that can be calculated (dynamic and static) and the appropriate perturbations used to calculate them.

3. Standard techniques for calculating response use either integer occupations (as appropriate for a system with an energy gap) or fractional occupations due to a smearing function, used to improve convergence for metallic systems. I will present a generalization which can be used to compute response for a system with arbitrary fractional occupations.

4. Chloroform (CHCl3) is a small molecule commonly used as a solvent in measurements of nonlinear optics. I computed its hyperpolarizability for second-harmonic generation with TDDFT with a real-space grid, finding good agreement with calculations using localized bases and with experimental measurements, and that the response is very long-ranged in space.

5. N@C60 is an endohedral fullerene, a sphere of carbon containing a single N atom inside, which is weakly coupled electronically. I show with TDDFT calculations that a laser pulse can excite the vibrational mode of this N atom, transiently turning on and off the system's ability to undergo second-harmonic generation. The calculated susceptibility is as large as some commercially used frequency-doubling materials.

6. A crucial question in understanding experimental measurements of nonlinear optics and their relation to device performance is the effect of the solution environment on the properties of the isolated molecules. I will consider possible explanations for the large enhancement of the hyperpolarizability of chloroform in solution, demonstrate an ab initio method of calculating electrostatic effects with local-field factors, and derive the equations necessary for a full calculation of liquid chloroform.

7. Many-body perturbation theory, in the GW approximation for quasiparticle bandstructure and Bethe-Salpeter equation for optical properties, is a powerful method for calculations in solids, nanostructures, and molecules. The BerkeleyGW code is a freely available implementation of this methodology which has been extensively tested and efficiently parallelized for use on large systems.

8. Molecular junctions, in which a single molecule is contacted to two metallic leads, are interesting systems for studying nanoscale transport. I will present a method called DFT+Sigma which approximates many-body perturbation theory to enable accurate and efficient calculations of the conductance of these systems.

9. Azobenzene is a molecule with the unusual property that it can switch reversible between two different geometries, cis and trans, upon absorption of light. I have calculated the structures of these two forms when absorbed on the Au(111) surface, to understand scanning tunneling microscope studies and elucidate the switching mechanism on the surface. I have also calculated the conductance of the two forms in a molecular junction.

10. The Seebeck and Peltier thermoelectric effects can interconvert electricity and heat, and are parametrized by the Seebeck coefficient. Standard methods in quantum transport for computing this quantity are problematic numerically. I will show this fact in a simple model and derive a more robust and efficient approach.

11. Pentacene is an organic semiconductor which shows exciton self-trapping in its optical spectra. I will present a method for calculation of excited-state forces with the Bethe-Salpeter equation that can be applied to study the geometrical relaxation that occurs upon absorption of light by pentacene.