Thermodynamic Modeling of Binary and Ternary Systems of Interest to Gas Antisolvent Precipitation

Show full item record

Title: Thermodynamic Modeling of Binary and Ternary Systems of Interest to Gas Antisolvent Precipitation
Author: McDiffett, Dana Marie
Advisors: Dr. George W. Roberts, Committee Member
Dr. Ruben G. Carbonell, Committee Co-Chair
Dr. Peter K. Kilpatrick, Committee Chair
Abstract: CO2 is a poor solvent for polar and complex molecules. The batch process known as GAS precipitation exploits the ability of CO2 to function as an antisolvent. CO2 is dissolved at low temperatures and moderate pressures into an organic solvent into which a solute has been dissolved. The liquid phase is expanded, reducing the solvent power, and causing precipitation of the solute. In this way, thermally-labile pharmaceutical compounds can be purified from organic solutions as very fine, uniform particles. In a simple (1) antisolvent — (2) solvent — (3) solute GAS system, solid-vapor-liquid equilibrium (SVLE) exists at the point of precipitation. The solubilities of acetaminophen in CO2-expanded ethanol, x3, and of acetaminophen in the equilibrium vapor phase, y3, were predicted using a three-component, ternary-phase Mathcad code. The Peng-Robinson equation of state was used to represent the fluid phases and a simple fugacity expression was used to represent the solid phase. Binary interaction parameters for the CO2-ethanol and ethanol-acetaminophen pairs, k12=0.0890 and k23=0.0099, were regressed from literature solubility data. The binary interaction parameter for the CO2-acetaminophen pair, k13=0.2614, was regressed using solubility data taken using a static equilibrium apparatus. The vapor phase mole fraction of acetaminophen in CO2 at 323K ranges from 10-6 to 10-4 over the pressure range 1500 psi to 4000 psi. The Peng-Robinson equation of state model is unable to capture the solubility behavior, but the data are adequately fit using a non-predictive density-based correlation. A sensitivity analysis conducted on the SVLE model to determine the effect of the binary interaction parameters on the prediction of x3 indicates that a poor fit for k13 does not prevent making an adequate prediction for the change in x3 with increasing pressure. At a given temperature, the solubility of acetaminophen in CO2-expanded ethanol decreases with pressure, indicating that GAS precipitation could be a viable means of forming acetaminophen particles. At a given pressure, the liquid phase acetaminophen solubility increases with temperature, indicating that higher operating temperatures will require higher operating pressures to achieve maximum solute product yields. In the future, experimental x3 data can be used to fit the k13 parameter in order to improve and validate the thermodynamic model.
Date: 2004-09-09
Degree: MS
Discipline: Chemical Engineering

Files in this item

Files Size Format View
etd.pdf 1.344Mb PDF View/Open

This item appears in the following Collection(s)

Show full item record