You are here:
COMPUTATIONAL CHEMISTRY METHOD FOR PREDICTING VAPOR PRESSURES AND ACTIVITY COEFFICIENTS OF POLAR ORGANIC OXYGENATES IN PM2.5
Citation:
Edney, E O., S. L. Clegg, L. J. Bartolotti, AND T. E. Kleindienst. COMPUTATIONAL CHEMISTRY METHOD FOR PREDICTING VAPOR PRESSURES AND ACTIVITY COEFFICIENTS OF POLAR ORGANIC OXYGENATES IN PM2.5. Presented at American Association for Aerosol Research, Charlotte, NC, October 7-11, 2002.
Impact/Purpose:
1. Using laboratory and field study data generated during FY99-FY04, develop a science version of a PM chemistry model for predicting ambient concentrations of water, inorganics, and organics in PM2.5 samples. The model will include the Aerosol Inorganic Model for predicting concentrations of inorganic compounds and a computational chemistry-based method for predicting concentrations of organic compounds.
2. Identify and evaluate methods for analyzing the polar fraction of PM2.5 samples.
3. Carry out short term field studies in Research Triangle Park, North Carolina in the summer and the winter to determine the composition of the organic fraction of ambient PM2.5 samples, with special emphasis placed on identifying and determining ambient concentrations of polar compounds.
4. Conduct laboratory studies to establish the chemical composition of secondary organic aerosol (SOA) and to determine source signatures for aromatic and biogenic SOA.
5. Conduct laboratory and theoretical investigations of thermodynamic properties of polar organic compounds.
6. Evaluate the science version of the PM chemistry model using laboratory and field data generated under this task as well as other available data in the literature.
7. Conduct PM chemistry-related special studies for OAQPS
Description:
Parameterizations of interactions of polar multifunctional organic oxygenates in PM2.5 must be included in aerosol chemistry models for evaluating control strategies for reducing ambient concentrations of PM2.5 compounds. Vapor pressures and activity coefficients of these compounds, some of which are semivolatile, are needed to predict the ambient concentrations of oxygenated compounds that may be distributed across a number of phases within the aerosol and in the gas phase. However, few measurements have been made to date of thermodynamic properties of these compounds, rather model values have been estimated mainly using predictive methods including UNIFAC that are based upon correlations between the molecular structure and thermodynamics properties. While UNIFAC has been a reliable tool in chemical engineering, its use in aerosol chemistry has been somewhat limited because the data sets upon which the models have been trained have not included extensive data for multifunctional oxygenates. In the present study, the computational chemistry-based model COSMO-RS is employed to predict the vapor pressures and activity coefficients of multifunctional oxygenated compounds. COSMO-RS is a quantum chemistry-based continuum solvation model where the molecular descriptors are (1) the electronic energies of the solute molecule in the gas phase and in an infinite conductor and (2) the surface screening charge distributions of the solute and solvent molecules in an infinite conductor. Statistical mechanics is then employed to calculate vapor pressures and activity coefficients. The molecular descriptors are calculated using quantum chemistry methods, while the statistical mechanics component requires a number of parameters that must be obtained by fitting the model to experimental data. However, the number of fitted variables required by COSMO-RS is far fewer than those needed by UNIFAC. Because COSMO-RS is based on first principles to a larger extent than UNIFAC, one might expect that it would be more valid for compounds whose structures have not been included in the training set. The purpose of this study is to investigate this hypothesis. Density functional theory is employed to calculate values for the COSMO-RS molecular descriptors for a number of oxygenated compounds whose experimental data are available. These data are used to obtain the COSMO-RS fitting parameters. Vapor pressures and activity coefficients for multifunctional oxygenates detected in PM2.5 samples, calculated using COSMO-RS, are compared with available experimental data.
This work has been funded fully, or in part, by the United States Environmental Protection Agency, under Contract Number 1D-5081-NANX to MCNC, Contract No. 68-D5-0049 to ManTech Environmental Technology, Inc and Contract 1D-5700-NATX to Dr. Simon L. Clegg. It has been subjected to Agency review and approved for publication.