Date of Award

2014

Document Type

Open Access Dissertation

Department

Chemistry and Biochemistry

Sub-Department

Chemistry

First Advisor

Sophya Garashchuk

Abstract

Molecular dynamics simulations, providing a detailed picture of the reaction mechanism, is an essential tool for theoretical and experimental chemists. In these simulations the nuclei are typically treated as classical particles, but under some conditions (low energies and temperatures, processes involving multiple electronic states) a classical description is inappropriate. Quantum effects of nuclear motion, such as tunneling and zero-point energy can play an important role in determining a reaction mechanism, yet exact quantum dynamics methods are limited to reactive systems of just 3-4 atoms. Central to this work is the development and implementations of an efficient trajectory-based methodology, in which the dominant quantum effects of nuclear motion are included through an approximate "quantum potential" term. A combination of quantum and classical nuclei can be evolved within this approach under the Hamiltonian or Boltzmann operators.

This quantum trajectory (QT) method is applied to the proton transfer in the enzymatic active site of soybean lipoxygenase-1. Experimental evidence suggests that this proton transfer step proceeds by a quantum tunneling mechanism. First, the reaction was examined as occurring within fixed substrate configurations at zero temperature, and the primary H/D kinetic isotope effect was in agreement with exact quantum and experimental results. Next, taking advantage of QT features, the effects of temperature and substrate motion were included into the simulation. Vibrational motion of the linoleic acid substrate was incorporated through on-the-fly density-functional tight-binding (DFTB) electronic structure (ES) calculations. This motion was found to modestly enhance the reaction across the temperatures of 250-350 K, and in a similar fashion for proton and deuteron. Through application of the quantum-mechanical flux operator and imaginary time evolution, the temperature was incorporated into the proton wavefunction. The experimentally observed weak temperature-dependence of the kinetic isotope effect was reproduced and is understood largely as an effect of the quantum partition function. Ideal scaling of the QTES-DFTB code with respect to the number of computing cores (typically run on thousands of cores), makes the developed methodology and code practical to chemical systems of up to 200 atoms.

Rights

© 2014, James William Mazzuca

Included in

Chemistry Commons

Share

COinS