Source: CONNECTICUT AGRICULTURAL EXPERIMENT STATION submitted to
ADSORBATE RETENTION BY SOIL ORGANIC MATTER: MOLECULAR SIMULATIONS AND EXPERIMENT
 
PROJECT DIRECTOR: Pignatello, J. J. Neimark, A. V.
 
PERFORMING ORGANIZATION
SOIL & WATER
CONNECTICUT AGRICULTURAL EXPERIMENT STATION
NEW HAVEN,CT 06504
 
NON TECHNICAL SUMMARY: We propose to characterize sorption by coupling molecular models and physical measurements. It is now possible to probe sorption at the molecular scale using Molecular Dynamics (MD) and Grand Canonical Monte Carlo (GCMC) methods, which are able to simulate both static and dynamic behavior. The compounds will include trichloromethane (grain fumigant), trichloroethylene (formerly used as a solvent carrier), dichlobenil (herbicide), and chloranil (fungicide). Four soils of various organic matter content are chosen. We will construct a computer model of a humic macromolecule and simulate its hydration with water. Macroscopic sorption/desorption experiments designed to compare with the simulation results will be conducted by well-established batch methods to obtain thermodynamic and kinetic parameters. The working hypothesis is that SOM is a three-dimensional phase that has properties similar to the glassy organic state. Sorption will be studied with respect to: prefered locations of molecules; the tendency of molecules to aggregate in voids; effects of sorption on void size and population; rates of diffusion; adsorption-desorption hysteresis; and competitive effects. This study will help assign mechanism of pesticide interaction, lead to breakthroughs in molecular modeling, and promote wider application of theoretical physics to environmental problems. The sorption of organic agrichemicals to soil organic matter (SOM) is fundamental to their bio-activity, leaching potential, and bioavailability, yet our understanding of the mechanism is incomplete.
 
OBJECTIVES: We propose to use computational molecular simulations to probe the mechanism of sorption of selected nonionic organic compounds to soil organic matter and to compare these simulations with experiment. The model compounds are trichloromethane, trichloroethylene, dichlobenil (2,6-dichlorobenzonitrile), and chloranil. Trichloromethane is an additive in commercial grain fumigant formulations; dichlorbenil is an herbicide; chloranil is a fungicide; trichloroethylene, no longer used in commercial pesticides, will be used here as a model compound. The first objective will be to construct a model of a humic macromolecule and simulate its interaction with water. Sorption will be then be investigated with respect to the following: preferential locations of adsorbate within the SOM psuedophase, the tendency of adsorbate molecules to aggregate in cavities within the organic matrix, effect of adsorbate sorption on cavity size and population, relative rates of diffusion within organic matter, adsorption-desorption hysteresis, and competitive effects in multiple adsorbate systems.
 
APPROACH: The first task will be to construct a model of a humic macromolecule, starting with a published suggested structure, and simulate its hydration with water. This stage will include choosing a suitable potential forcefield, calculation of partial charges assigned to each atom of SOM, optimization of the model SOM structures based on the forcefields chosen; and building potential energy surfaces for water. Our working hypothesis, based on much experimental (but macroscopic) evidence, is that SOM is a three-dimensional macromolecular phase that has properties akin to the glassy organic state. Macroscopic sorption/desorption experiments designed to compare with the simulation results will be conducted by well-established batch methods. Data-intense isotherms (>70 points) will be constructed on single and binary solute systems to obtain dual-mode parameters. Relative effective diffusivities for sorption and desorption will be measured based on dual-mode kinetic models. To arrive at an explanation for hysteresis, we will measure a thermodynamically-based hysteresis index for CHCl3 and TCE adsorption-desorption loops and compare them with the hysteresis index of CO2. The effect of pre-conditioning on sorption will attempt to confirm a hole enlargement hypothesis.
 
CRIS NUMBER: 0186017 SUBFILE: CRIS
PROJECT NUMBER: CONH00463 SPONSOR AGENCY: CSREES
PROJECT TYPE: NRI COMPETITIVE GRANT PROJECT STATUS: TERMINATED MULTI-STATE PROJECT NUMBER: (N/A)
START DATE: Dec 1, 2000 TERMINATION DATE: Nov 30, 2003

GRANT PROGRAM: SOILS & SOIL BIOLOGY
GRANT PROGRAM AREA: Natural Resources

CLASSIFICATION
Knowledge Area (KA)Subject (S)Science (F)Objective (G)Percent
133522020006.1100%

CLASSIFICATION HEADINGS
KA133 - Pollution Prevention and Mitigation
S5220 - Pesticides
F2000 - Chemistry
G6.1 - Ensure Clean Water and Air


RESEARCH EFFORT CATEGORIES
BASIC 90%
APPLIED 10%
DEVELOPMENTAL (N/A)%

KEYWORDS: soil organic matter; soil chemistry; sorption; pesticides; pollution control; dynamics; monte carlo method; hysteresis; humic acids; diffusion; kinetics; simulation models; mechanisms; sorption isotherms; size; rates; adsorption; desorption; molecular structure; hydration; optimization

PROGRESS: Dec 1, 2000 TO Nov 30, 2003
A three-dimensional molecular model was constructed based on elemental and group composition of humic acid (HA) and fulvic acid (FA). Molecular dynamics (MD) simulations at constant pressure were used to obtain equilibrium configurations and to estimate flexibilities of different model HA fragments for use in large-scale lattice Monte-Carlo simulations. To predict dissolution of organic compounds we employed UNIFAC-FV (free volume) group-contribution model. Calculated partition coefficients of typical contaminants in dry and hydrated model HA compared favorably with their octanol-water partition coefficients. Measured partition coefficients of probes (cyclohexane, benzene, methyl benzoate, acetophenone, and anisole) in Leonardite HA and HA extracted from a peaty soil are compared with calculated values. To study mechanisms of hole-filling, melting and glass transition of HA by Monte Carlo simulations, a coarse-grained lattice model was constructed based on the MD simulation results. Sorption of carbon dioxide was studied on soils, humin and HA fractions. Carbon dioxide sorption was strongly hysteretic, especially for soils with a high organic matter content. Sorption could be described as a two-stage process characterized by different short-time and long-time rates. The first stage (hours to several days) is controlled by diffusion through rubbery regions of SOM and adsorption in (desorption from) accessible micropores in glassy regions of SOM. A Diffusion-Controlled Hysteresis (DCH) model is introduced to extract equilibrium isotherms and diffusion coefficients from nonequilibrium hysteretic measurements. The shapes of the equilibrium isotherms unambiguously demonstrate the microporous nature of SOM. Second stage uptake or release scales linearly with the square root of time, and the characteristic time scale is much larger (months). Using the DCH model, we show that the long-time kinetics is correlated with the increasing sorption of CO2 in micropores. We hypothesize that the second stage is caused by a slow diffusion into remote structural domains, which become accessible via structural swelling and/or restructuring of the solid matrix. These domains are thought to be responsible for irreversible sorption and sequestration of organic molecules. Conditioning effect experiments are used to investigate history-dependent sorption behavior of organic solutes in natural organic matter solids. The conditioning effect was observed in a soil humic acid in the hydrogen ion form and in the same humic acid in the aluminum-ion exchanged form. The conditioning effect is consistent with the presence of deformable pores. The conditioning effect for 1,2,4-trichlorobenzene in a conditioned peat soil decayed with sample heating. The decay was biphasic and the rate constants increased exponentially with temperature. The results are tentatively interpreted in terms of a two-compartment (elastic and viscoelastic) model of matrix relaxation.

IMPACT: 2000-12-01 TO 2003-11-30 The objective of this project is to use molecular models and physical measurements to characterize sorption of hydrophobic organic contaminants (HOC) in soils. The short and long term hysteresis of CO2 is highly significant in that it suggests that sorption to soil organic matter is subject to continued slow changes over months even for small gaseous molecules. The hysteresis observed for CO2 and the conditioning effect observed for trichlorobenzene signifies the existence of deformable pores in humic substances and lends support to the glassy polymer paradigm for humic substances. The DCH model of sorption kinetics and corresponding experimental protocols constitute a basis for characterization of microporosity in soils. The theoretical work suggests UNIFAC-FV may be useful as predictive tool for sorption.

PUBLICATION INFORMATION: 2000-12-01 TO 2003-11-30
Ravikovitch, P.I., A. V. Neimark, W.J. Braida, and J.J. Pignatello, Sorption of Carbon Dioxide in Soils and Soil Organic Matter Fractions. Diffusion Controlled Hysteresis Model, Pore Structure Characterization, and Long Time Kinetics., Environ. Sci. Technol., 2004, under review.

PROJECT CONTACT INFORMATION
NAME: Pignatello, J. J.
PHONE: 203-974-8518
FAX: 203-974-8502