Open Access
Castro Marcano, Fidel
Graduate Program:
Materials Science and Engineering
Master of Science
Document Type:
Master Thesis
Date of Defense:
December 15, 2009
Committee Members:
  • Coray M Colina, Thesis Advisor
  • molecular modeling
  • polymer phase equilibria
  • Sanchez-Lacombe
  • gas adsorption
  • hydrogen storage
In this work, a parameterization strategy that allows the calculation of polymer molecular parameters from macroscopic properties of binary polymer solutions is presented. The proposed parameterization is demonstrated by reference to the PC-SAFT equation of state, but can be applied to any molecular-based model. The parameterization scheme has been developed in terms of the polymer-solvent interaction parameter and the Hildebrand parameter, which describe the molecular nature and extent of the polymer-solvent interactions. The specification of these macroscopic properties yields a set of polymer parameters that are suitable for the description of thermodynamic properties and phase behavior of polymer solutions. In this way neither extensive experimental data nor complex minimization techniques are necessary, as is required for the current approaches for the estimation of pure-polymer parameters for SAFT-type equations. Using polymer parameters calculated from the proposed parameterization strategy, the PC-SAFT model could satisfactorily predict the phase equilibria, gas solubility and polymer swelling behavior of binary and ternary polymer solutions with different solvents, including nonassociating compounds such as n-alkanes, polar compounds such as ethers, esters and ketones, and associating compounds such as alcohols. A computational approach for building atomistic models for amorphous polymer networks in order to simulate their pore structure and gas adsorption properties is also presented. The computational approach replicates the basic reactivity rules of the self-condensation reaction of dichloroxylene (DCX) via Friedel-Crafts chemistry and allows the formation of amorphous polymer networks, which are not possible to generate by structural X-ray crystallography/diffraction as is usually done for crystalline materials. The method is discussed for poly(dichloroxylene) networks, but can be extended to other polymer networks. Atomistic models were further refined by fitting to characterization data (i.e., bulk density, absolute density, micropore volume and elemental composition). These models were characterized by specific surface area and pore size distribution. A sensitivity analysis was performed to determine the minimum box size that should be used in adsorption simulations. Simulated adsorption isotherms and isosteric heats for methane and hydrogen were found to be in reasonable agreement with the experimental data.