Environ. Sci. Technol. 2001, 35, 2518-2524
Solvation Effect on Organic Compound Interactions in Soil Organic Matter MIKHAIL BORISOVER,* MINOLEN REDDY, AND ELLEN R. GRABER Institute of Soil, Water and Environmental Sciences, The Volcani Center ARO, P.O. Box 6, Bet Dagan, 50250 Israel
We examine sorption of pyridine by soil organic matter (SOM) from different organic media including n-hexadecane, acetonitrile, acetone, and n-hexadecane mixtures with either acetonitrile or acetone and compare it with sorption from water. By using an activity-based comparison, we distinguish between solvent-assisted and solvent-competitive sorption behavior. Pyridine was selected because it forms strong complexes with phenolic and carboxylic groups, such that site interactions should dominate interactions in SOM. It is anticipated that pyridine sorption will be illustrative of the importance of disrupting strong interactions in a condensed, shrunken SOM phase for many organic compounds. It was generally found that activitynormalized pyridine uptake was assisted by polar solvent molecules rather than suppressed due to competition. An explanation is tendered on the basis of our earlier hypothesis of water-assisted disruption of polar SOM contacts. Certain polar moieties of dry SOM are unavailable for compound sorption due to strong interactions between them. By penetrating SOM structure, solvent molecules (and water) solvate (hydrate) polar moieties creating new sorption sites. Solvent molecules must solvate both moieties of the polar contact, such that the driving force for solventassisted sorption is solvation of the partner of the disrupted contact that does not directly interact with the sorbate.
Introduction The hydration status of soil organic matter (SOM) is wellknown to affect its structure. Multiple processes including hydrogen bonding between polar functional groups, conformation rearrangements, and bridging via metal cations may occur in SOM humic substances when partially or completely dehydrated (1-4). These processes cause the macromolecular substances to shrink, creating a more condensed structure in the dehydrated state. It is expected that such a condensed structure may have an increased thermodynamic and/or kinetic potential to resist sorption of organic compounds due to diminished availability of sorption sites and reduced sorbate diffusivity. Shrinkage of the SOM structure may strongly affect desorption of organic compounds, contributing to decreased rates of organic sorbate release under dry or wet conditions (4). Comparing sorption of selected organic compounds on dry and hydrated SOM-rich sorbents is a means of evaluating the effect of hydration on sorption and can elucidate certain * Corresponding author telephone: 972-3-968-3314; fax: 972-3960-4017; e-mail:
[email protected]. 2518
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 12, 2001
properties and structures of condensed SOM. An increase in vapor-phase sorption of trichloroethylene on Aldrich humic acid salt at an intermediate water humidity (5) demonstrates that interactions between polar SOM moieties have an important role in limiting sorption in dry SOM. A strong increase in activity-normalized sorption of the specifically interacting compounds pyridine and phenol on SOM from water as compared with n-hexadecane was also reported (6). This hydration-assisted sorption was interpreted in terms of disruption of SOM polar contacts in the presence of water. A weak reduction in sorption of benzene, carbon tetrachloride, and trichloroethylene by SOM-rich sorbents from water as compared with vapor-phase sorption on dry SOM (7) shows that, for these compounds, complete hydration of the SOM sorbent did not result in exposure of new effective sorption sites. Complete hydration of SOM involves dissociation of carboxylic groups and strong hydration of many other polar functionalities, thus disrupting SOM structure. We hypothesized that by measuring sorption of a selected organic compound on a SOM-rich sorbent from different organic media ranging from nonpolar to polar, different extents of solvent-disrupting effects could be tested. On the basis of such experiments, it would be possible to distinguish between cases where the solvent may assist in activity-normalized sorption of the probe compound (solvent-assisted sorption) and cases where the solvent would successfully compete with the probe compound for sorption sites (solvent-competitive sorption). This comparison could be made more sophisticated by measuring sorption from n-hexadecane mixtures with chosen polar solvents, making it possible to observe the gradual exposure of new sites for SOM-sorbate interactions. By comparing the hydration effect with the organic solvent effect, it could be possible to elaborate on the role of hydration in sorption. Such a solvent-based approach should not be confused with studies of sorption on soil from aqueous solutions of miscible organic solvents (e.g., refs 8-10) or studies of sorption on soil from nonaqueous solvents (e.g., refs 11 and 12). In experiments with water-miscible solvents, the effect of the solvent on SOM structure can only be examined in relation to a strongly hydrated SOM structure and thus cannot elucidate organic solvent or water effects on a dehydrated SOM structure. In sorption experiments from nonaqueous solvents on mineral-rich soils, the strong sorption potential of dry mineral surfaces interferes with evaluating SOMorganic compound interactions. Sorption data for organic compounds on SOM-rich sorbents from organic solvents are scarce and limited mainly to sorption isotherms measured from hydrocarbons (4, 6, 13, 14). Apparently, there is no activity-based comparison of organic compound sorption on SOM-rich sorbents from organic media with different solvation abilities. As such, we examine sorption of a selected specifically interacting compound, pyridine, by a model soil organic matter from different organic media including n-hexadecane, acetonitrile, acetone, and n-hexadecane mixtures with either acetonitrile or acetone and compare it with sorption from water. By using an activity-based comparison of sorption data, we intend to distinguish between solvent-assisted and solvent-competitive sorption behavior. Pyridine was selected according to its ability to undergo specific interactions with SOM (15). On the basis of gas-phase-SOM distribution coefficients (15), it may be calculated that pyridine interacts with hydrated SOM 2500 times stronger than its nonpolar analogue, benzene. Pyridine is capable of forming strong 10.1021/es001810d CCC: $20.00
2001 American Chemical Society Published on Web 05/11/2001
TABLE 1. Experimental Conditions for Sorption Experiments Carried Out in Different Media
expt series I II III IV V VI VII
solute/solvent pyridine/n-hexadecane pyridine/acetonitrile pyridine/acetone acetonitrile/n-hexadecane pyridine/n-hexadecane + acetonitrile (+pyridine) acetonitrile/n-hexadecane + acetonitrile (+pyridine) pyridine/n-hexadecane + acetone (+pyridine) acetone/n-hexadecane (+pyridine) acetonitrile/water
solid:liquid sorption initial solute ratio time (h) concn rangea 1:30 1:10 1:3 1:25 1:30 1:30 1:6 1:6 1:2, 1:3
940c 900 1000 2060 2300 2300 1600 1600 16
27-207 50-7500 68-5600 0.018-0.9 58-600 0.05-0.9 100-600 5-15 5
initial concna at which sorption kinetics were followed
sorbed fraction of total amount (%)
Kdb range (mL/g)
nad 230 200 0.14, 0.36, 0.72 58, 400, 600 0.05, 0.2, 0.5 100, 600 5, 10, 15 na
17-40 15-50 10-50 30-70 20-80 25-70 60-80 ∼20 0
6-20 1.5-9 0.3-3 10-50 7-140 10-70 8-25 ∼1.6 0
a In mg/L for pyridine or % v/v for acetonitrile and acetone. b Distribution coefficient K is defined as ratio of sorbed concentration to solution d concentration. c Apparent sorption equilibrium is expected at 400 h (6). d na, not applicable.
hydrogen bonds and proton-transfer complexes with phenolic and carboxylic groups of organic compounds (16, 17), such that site interactions should dominate pyridine interactions in the SOM phase. If dehydrated SOM resists pyridine sorption, such resistance would be even more pronounced for many other organic compounds less capable of specific interactions. It is thus anticipated that pyridine sorption will be illustrative of the importance of disrupting strong interactions in a condensed, shrunken SOM phase for many organic compounds. Implications concerning the hydration effect on sorption of organic compounds by SOM will be derived from sorption data.
Experimental Section Materials. Pyridine (chromatographic grade puriss. p.a. >99.8%) from Fluka, n-hexadecane (99+%) from Aldrich, acetonitrile (analytical grade) and acetone (analytical grade) from Bio-Lab, and water (double distilled) were used without additional purification. Pahokee peat purchased from the International Humic Substances Society (83% organic matter) was freeze-dried and used as a model organic matter-rich sorbent. The peat sample contained 49% C, 3.3% N, 4.3% H, and 0.5-1.2% S on a dry weight basis as determined by elemental analysis (Carlo Erba, EA-1108). Moisture content of the freeze-dried peat was in the range of 2-3% w/w as determined by oven-drying at 105 °C. Sorption from Organic Solvents. Details of kinetic and equilibrium sorption experiments are summarized in Table 1. For evaluating sorption kinetics, peat suspensions in prepared solutions were mixed by head-over-end shaker continuously at 25 ( 1 °C in the dark in glass vials sealed by Teflon mininert valves. After being centrifuged, 1 µL of supernatant was obtained through the mininert valve and analyzed by GC/FID (Varian 3300, manual on-column injection, J&W DB-1 column, 30 m, 0.53 mm i.d., 1.5 µm film.). Each system was allowed to reach apparent equilibrium, generally within 700-800 h. Sorption experiments were duplicated or triplicated. For obtaining equilibrium sorption data, suspensions were mixed under the same conditions in glass vials sealed by screw caps equipped with Teflon-lined silicone septa. After apparent sorption equilibrium was reached, all samples (kinetic and equilibrium series) were centrifuged, and solute concentration in the solution phase was determined. Solute losses in blank solutions were small (pyridine in n-hexadecane
