With advancements in computer technology and processing power, the ability to examine chemical systems using theory continues to be more practicable. Using ab initio methods, such as density functional theory, we are now able to routinely simulate hundreds of atoms. This system size allows us to directly simulate surfaces and nano-materials that are industrially relevant. With the expansion of accessible systems comes the opportunity to develop new computational methods to extract their chemical properties.
Of particular interest is bridging the time scale gap between simulation and experiment. The evolution of a system chemical in time can be directly simulated using classical dynamics, however, molecules vibrate on the order of femtoseconds and interesting transitions tend to happen on much longer time scales: milliseconds to seconds. In condensed phase chemical systems these interesting transitions are hindered by energy barriers so state to state dynamics are dominated by rare evens. Luckily, rare event transitions tend to happen through mountain passes in the potential energy landscape. Within harmonic transition state theory, the transition states between minima can be characterized by saddle points. Finding saddle points is a challenging problem which has not been satisfactorily solved; nevertheless, there are algorithms currently being used despite their deficiency. In particular, my work strives to improve the efficiency and stability of the nudged elastic band method and compare its performance to similar algorithms on a variety of test systems.
In addition, I present a method to predict how energy-based chemical properties change with respect to the chemical composition of the system. This is achieved by taking a derivative of the property with respect to the atomic numbers of the atoms present in the system. The accuracy and predictive quality of these derivatives are assessed for both model and industrially relevant systems. With this information, we can follow these derivatives to optimize a desired property in the space of chemical composition. This method is a step toward using theory to rationally design compounds with desirable properties. / text
| Identifer | oai:union.ndltd.org:UTEXAS/oai:repositories.lib.utexas.edu:2152/ETD-UT-2010-12-2179 |
| Date | 07 February 2011 |
| Creators | Sheppard, Daniel Glen |
| Source Sets | University of Texas |
| Language | English |
| Detected Language | English |
| Type | thesis |
| Format | application/pdf |
Page generated in 0.002 seconds