A computer model of water-in-oil emulsion coagulation

• Main Author: Seymour, Lisa
• Other Authors: Jackson, Graham Ellis
• Format: Doctoral Thesis
• Language: English
• Published: University of Cape Town 2016
• Subjects: Chemistry
• Online Access: http://hdl.handle.net/11427/17966

Include bibliographical references. === In this thesis, a stochastic computer model of water-in-oil emulsion coagulation, a two stage process of aggregation and coalescence, is presented. The theoretical basis of the model, including equations for the van der Waals, electrostatic and steric energy barriers between dissimilar droplets, is described. Many of these equations have been derived by the author. A chemical speciation study of the aqueous phase typically found in emulsion explosives is presented. A potentiometric investigation of the protonation equilibria of propionate, succinate and mono-methyl succinate in tetraethyl ammonium bromide, ammonium nitrate, sodium nitrate, potassium nitrate and calcium nitrate at 25°C and 3 mol/dm³ ionic strength was performed. Nuclear Magnetic Resonance titrations for succinate and propionate in varying concentrations of the same salts are also shown. A method of converting thermodynamic stability constants from one ionic strength to another using a modified form of the Pitzer equations is presented with a computer program which performs the conversion. A novel method of obtaining complexation constants from protonation constants in varying media is proposed. Using optical microscopy, creaming rates and laser particle sizing, the affects of changing surfactant concentration, salt concentration, pH and shearing time for emulsions of ammonium nitrate solution in heptane with CRILL 43 are shown. Equations are derived for converting creaming rate data to droplet size information and a computer program for converting Malvern light intensity data in the anomalous regime (typical of water-in-oil emulsions) to size distribution data is presented. The computer model is validated against experimental data from this work and the literature and is used to make stability predictions for systems for which no data exists. Further uses for the model are discussed.