Font Size: a A A

Statistical mechanical and quantum mechanical modeling of condensed phase systems

Posted on:2010-03-30Degree:Ph.DType:Dissertation
University:University of PittsburghCandidate:LaBrosse, Matthew RFull Text:PDF
GTID:1440390002490134Subject:Chemical Engineering
Abstract/Summary:
Understanding adsorption in nanoporous media is vital to improving their use in industrial applications such as fluid storage and separations processes. One major objective of this research is to shed light on an on-going controversy in literature over where gases adsorb on single walled carbon nanotube bundles. Grand-canonical Monte Carlo simulations have been performed using models of carbon nanotube bundles composed of tubes of all the same diameter (homogeneous) and tubes of different diameters (heterogeneous). We used three metrics with which we compared our simulation results to those found in experiments on carbon nanotubes: the specific surface area, the isosteric heat of adsorption, and the adsorption capacity. Simulations of classically behaved fluids Ar, CH4, and Xe indicate that nanotubes prepared by the HiPco process are best described by a heterogeneous bundle model with ∼11% of the nanotubes opened. Ne gas requires additional considerations to describe the quantum effects at the temperatures of interest, which have been implemented by the Feynman-Hibbs approximation. Overall, calculated results from Ne simulations are consistent with those from classical fluids. However, Ne simulations strongly indicate that the small interstitial channels formed by exactly three nanotubes are closed. Combined with previous studies on classically behaved fluids Ar, CH4, and Xe, experimental data including Ne are best matched by hetergeneous bundles with ∼11% open-ended nanotubes.;The development of a heterogeneous Co/C/O reactive force field (ReaxFF) potential has also been a major objective of this research. ReaxFF provides a method to describe bond-breaking and bond-forming events that can be applied to large-scale molecular dynamics (MD) simulations. This many-bodied semi-empirical potential has been trained from ab initio density functional theory (DFT) calculations. The training set originally included descriptions of bulk and surface condensed phase cobalt systems. This was later expanded to include binary (Co/C, Co/O) and tertiary (Co/C/O) heterogeneous interactions. We have tested these parameters against additional DFT calculations not included in the training set. The parameter optimization has produced a force field capable of describing additional configurations with the same accuracy as those used in the fitting procedure. The optimized parameters have been used to predict the melting point and diffusion coefficients of condensed phase cobalt. Large-scale simulations of a Co2C phase nanoparticle show segregation on short time scales (less than 300 ps), with all C atoms forming chains and small graphene structures on the surface of a solid Co nanoparticle core. ReaxFF has also been used to show that diffusion of Co is more energetically favorable than oxygen through the interstitial sites of a cobalt oxide crystal. This is consistent with experimental observations that oxidized cobalt nanoparticle form hollow cobalt oxide nanospheres due to a faster Co diffusion rate through the oxide layer. These two binary applications demonstrate that ReaxFF is transferable to heterogeneous systems and is a computationally inexpensive means by which transition metal surface reactions can be explored.
Keywords/Search Tags:Condensed phase, Heterogeneous, Surface
Related items