We have modelled structures and dynamics of hydrogen bond networks that form from imidazoles tethered to oligomeric aliphatic backbones in crystalline and glassy phases. We have studied the behavior of oligomers containing 5 or 10 imidazole groups. These systems have been simulated over the range 100-900 K with constantpressure molecular dynamics using the AMBER 94 force field, which was found to show good agreement with ab initio calculations on hydrogen bond strengths and imidazole rotational barriers. Hypothetical crystalline solids formed from packed 5-mers and 10-mers melt above 600 K, then form glassy solids upon cooling. Viewing hydrogen bond networks as clusters, we gathered statistics on cluster sizes and percolating pathways as a function of temperature, for comparison with the same quantities extracted from neat imidazole liquid. We have found that, at a given temperature, the glass composed of imidazole 5-mers shows the same hydrogen bond mean cluster size as that from the 10-mer glass, and that this size is consistently larger than that in liquid imidazole. Hydrogen bond clusters were found to percolate across the simulation cell for all glassy and crystalline solids, but not for any imidazole liquid. The apparent activation energy associated with hydrogen bond lifetimes in these glasses (9.3 kJ/mol) is close to that for the liquid (8.7 kJ/mol), but is substantially less than that in the crystalline solid (13.3 kJ/mol). These results indicate that glassy oligomeric solids show a promising mixture of extended hydrogen bond clusters and liquid-like dynamics. This study prompted a continued look at smaller oligomers (monomers, dimers, trimers, and pentamers). Using many of the above statistics we found that decreased chain length decreased the tendency to form global hydrogen bonding networks (percolation pathways). We also developed an reorientational correlation for the imidazole ring which allowed us to extract a timescale for reorientation. Smaller chains produce faster reorientation timescales and thus there is a trade off between faster reorientation dynamics and long global hydrogen bonding networks. Moreover we showed that homogeneity of chain length has no effect on hydrogen bonding statistics. Initial development on a multi-state empirical valence bond model has been to study proton transfer in liquid imidazole. We have shown that GAFF produces very large proton transfer barriers created by a highly repulsive N˙ ˙ ˙H VDW interaction at the transition point. In order to produce an acceptable fit to the potential energy surface while still producing stable dynamics this interaction must be turned off. This is in contrary to what is reported in the literature [14]. Using our model we have produced simulations with acceptable drift in the total energy (3.2 kcal/mol per ns) and negligible drift in the temperature (.12 K/ns).
| Identifer | oai:union.ndltd.org:UMASS/oai:scholarworks.umass.edu:dissertations-7059 |
| Date | 01 January 2013 |
| Creators | Harvey, Jacob A |
| Publisher | ScholarWorks@UMass Amherst |
| Source Sets | University of Massachusetts, Amherst |
| Language | English |
| Detected Language | English |
| Type | text |
| Source | Doctoral Dissertations Available from Proquest |
Page generated in 0.002 seconds