Polymerization of Vinylidene Fluoride in Supercritical Carbon Dioxide: Molecualr Weight Distribution

Show full item record

Title: Polymerization of Vinylidene Fluoride in Supercritical Carbon Dioxide: Molecualr Weight Distribution
Author: Saraf, Manish Kumar
Advisors: Dr. George W. Roberts, Chair
Dr. Joseph M. DeSimone, Co-Chair
Dr. Saad A. Khan, Member
Abstract: Conventional vinylidene fluoride polymerizations require the use of large amounts of water. Therefore, energy and cost-intensive drying and separation steps are required to isolate the polymer. Additionally, some conventional polymerizations use fluorosurfactants that belong to 3M's 'Scotchgard' family of surfactants, which are being phased out due to environmental concerns. In this research, we investigate a "green" process for the continuous polymerization of vinylidene fluoride by free-radical precipitation polymerization in supercritical carbon dioxide (scCO2). Significant technological and environmental improvements can be achieved by this technology, such as: 1) elimination of waste streams generated by conventional suspension, and emulsion processes; and 2) achievement of major energy savings, as the polymer is isolated in a dry form with no water or solvent to evaporate. The experimental system used in this research consisted of a continuous stirred tank reactor (CSTR) for polymerization, and a polymer withdrawal system where polymer was collected, and continuously ejected to ambient conditions. The polymer was collected as a dry "free-flowing" powder, and was characterized by GPC (gel permeation chromatography), DSC (differential scanning calorimetry), and Fluorine-19 NMR (nuclear magnetic resonance spectroscopy). Experiments were carried out to study the effect of different parameters such as inlet monomer concentration ([M]in= 0.78 to 3.5 M), pressure (P=3050 to 4400 psig) , temperature (T=65 to 85 oC) , and average residence time (t = 12 to 50 mins.), on the polymerization rate and the average molecular weights. A homogeneous model based on classical free radical kinetics predicted the polymerization rate very well at the lower rates. However, this could not predict an inhibition in the rate that was observed at higher monomer concentrations. A bimodal molecular weight distribution was obtained at inlet monomer concentrations greater than around 1.9 M. The second (high molecular weight) mode was more prominent at higher monomer concentrations, higher residence times, and lower temperatures. Two hypothesis, poor mixing and long chain branching, were investigated to explain these broad and bimodal molecular weight distributions (MWDs). Agitation studies showed that poor mixing can account for the slower rate of polymerization at high monomer concentrations, but does not to bimodal MWDs. End group analysis using NMR showed that chain transfer occurs, probably to polymer. A homogeneous kinetic model was developed to investigate the effect of chain branching arising from chain transfer to polymer. This model showed that broad MWDs with large polydispersity indices (PDIs) are obtained at high monomer concentrations and at high residence times. These predictions matched very well with experimental data. However, the model predicted a unimodal MWD even at very high values of the rate constant for chain transfer to polymer.Batch polymerization studies were carried out to develop alternate low temperature initiators, which could reduce operating and equipment costs and potentially produce high molecular weight polymer with reasonable yields. A fluorinated initiator obtained from hexafluoropropynl oxide dimer produced PVDF with very high molecular weights (~90 K) at reasonable yields.
Date: 2001-11-20
Degree: MS
Discipline: Chemical Engineering
URI: http://www.lib.ncsu.edu/resolver/1840.16/2179

Files in this item

Files Size Format View
etd.pdf 750.5Kb PDF View/Open

This item appears in the following Collection(s)

Show full item record