1999 Progress Report: In situ Diagnostic Techniques for Probing Solvation Effects in Supercritical Fluid Reaction Media for Synthetic Organic ChemistryEPA Grant Number: R826738
Title: In situ Diagnostic Techniques for Probing Solvation Effects in Supercritical Fluid Reaction Media for Synthetic Organic Chemistry
Investigators: Steinfeld, Jeffrey I. , Tester, Jefferson W.
Institution: Massachusetts Institute of Technology
EPA Project Officer: Hahn, Intaek
Project Period: October 1, 1998 through September 30, 2001
Project Period Covered by this Report: October 1, 1998 through September 30, 1999
Project Amount: $265,000
RFA: Technology for a Sustainable Environment (1998) RFA Text | Recipients Lists
Research Category: Sustainability , Pollution Prevention/Sustainable Development
Objective:The use of supercritical fluids (SCFs) as nontoxic solvent replacements in fine chemical synthesis is one strategy for waste minimization that is gaining increasing attention in research and industrial applications. Because SCF properties change significantly with relatively small changes in pressure or density in the critical region, parameters such as solubilities, reaction rates, and selectivities may be "tunable," making SCFs a particularly versatile and desirable reaction media. However, basic data and theoretical models are lacking for solvation of reactants and the influence of solvent density on reaction pathways in SCFs. This is due in part to the difficulty of making measurements in these high-pressure, sometimes high-temperature fluids. Raman spectroscopy has been demonstrated as a feasible in situ, noninvasive measurement technique in SCFs, but its full potential has not yet been realized.
In this research program, we plan to: (1) develop Raman spectroscopy as a robust, in situ diagnostic technique to address these fundamental questions; and (2) provide predictive modeling tools to enhance commercial applications of SCF reaction chemistry. Raman spectroscopy will be used to probe local solvation effects on species dissolved in supercritical carbon dioxide and near-critical water. Using band shape analysis to determine local densities, we can quantify the effects of density on reaction rates and pathways. Local solvent densities measured in SCFs will be used to determine empirically correct intermolecular potential functions, which will enable molecular modeling tools to predict solubilities and reaction dynamics in these fluids. This molecular-level understanding is critical to designing chemical processes using SCF solvents.
Progress Summary:In the first year of our project, significant progress was made towards better understanding the solvation of species in supercritical fluids and the effects these solvents have on reactions. Progress has been made in both the development of Raman spectroscopy for measurement of solvation effects in supercritical CO2 (scCO2) systems and the modeling of solvent effects on reactions in near-critical water systems.
Raman Spectroscopy. In the past year, a Raman spectrometer was donated to the Massachusetts Institute of Technology (M.I.T.) by Asahi Chemical Company and assembled for use with supercritical systems. A small high-pressure cell was designed and built with sapphire windows to allow for in situ measurements. The view cell currently is capable of temperatures up to 100 C and pressures up to 150 bar, which is sufficient for a wide range of densities of CO2 (Tc = 31.1 C, Pc = 74 bar). However, modifications of the seals are needed to reach the conditions necessary for near-critical or supercritical water (Tc = 374 C, Pc = 221 bar).
The scCO2 cell was used to measure the Fermi dyad of the symmetric stretch of pure CO2 as the temperature and pressure was varied. A slight red-shift of the bands was observed as the density was increased to liquid-like conditions from the original gaseous spectrum. No dependence on temperature was observed for similar density conditions.
Measurements also were conducted on benzene (C6H6) and methylene chloride (CH2Cl2) dissolved in CO2. For each of these species, two vibrational modes were measured as the conditions of the solvent were varied from gas-like conditions to liquid-like conditions. A red-shift was observed for all of the vibrational modes as the density increased and was most significant for the symmetric C-H stretch mode of benzene (n1). For similar densities, lower temperatures resulted in a more significant red-shift.
Qualitatively, we can relate the red-shift of the vibration to the degree of solvation of the solute by CO2 because the solutes are stabilized by the solvent. By this same argument, one can explain the dependence on temperature by noting that for similar densities, it is more difficult for a solvent to align itself to stabilize the solute at higher temperatures. Combining these measurements with molecular dynamics will allow a quantitative determination of the local density and solvation of molecules in scCO2.
Effect on Reactions in Near-Critical Water. Local solvation effects have a significant effect on the reaction rate of hydrolysis reactions in near-critical water. Experiments have been performed on the rate of methyl t-butyl ether (MTBE) hydrolysis and decomposition in water over the temperature range of 200-600 C and a pressure of 250 bar.
(CH3)3COCH3 --> H2O a CH3OH + (CH3)3OH
As the temperature is increased (at constant pressure), the rate of reaction reaches a local maximum just below the critical temperature, followed by a significant drop at the critical point to a local minimum. This unique rate behavior is due to the dramatic changes in the dielectric strength of water. Water changes from a very polar solvent at ambient temperature (e = 79 @ T = 25 C), to a moderately polar solvent below the critical point (e = 11 @ T = 370 C), and to a nonpolar solvent above the critical point (e = 2.5 @ T = 400 C). Because the transition state of the reaction has a larger dipole moment than the MTBE, the polarity of the solvent stabilizes the transition state relative to the reactant, effectively decreasing the activation barrier for more polar solvents.
Ab initio calculations were performed using Gaussian 98 to determine the transition state of the unimolecular decomposition reaction. The solvent was modeled using the isodensity polarized continuum model, and energy calculations were performed using density functional theory (B3LYP). The calculations showed the relative decrease of the activation energy with increasing dielectric strength and the local maximum below the critical point; however, the calculations also showed that the unimolecular decomposition reaction is small relative to the total rate. This agrees with the observation of t-butanol appearing early in the experimental measurements. Efforts currently are focused on determining the transition state of the hydrolysis reaction.