JEQ Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (4)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zhu, D.
Right arrow Articles by Schlautman, M. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zhu, D.
Right arrow Articles by Schlautman, M. A.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Zhu, D.
Right arrow Articles by Schlautman, M. A.
Related Collections
Right arrow Isotopes
Right arrow Organic Compounds
Right arrow Soil Pollution
Right arrow Water Pollution
Journal of Environmental Quality 32:232-239 (2003)
© 2003 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America

TECHNICAL REPORTS
Organic Compounds in the Environment

Molecular-Level Investigation of Monoaromatic Compound Sorption to Suspended Soil Particles by Deuterium Nuclear Magnetic Resonance

Dongqiang Zhu*,a, Bruce E. Herberta and Mark A. Schlautmanb

a Dep. of Geology and Geophysics, Texas A&M Univ., College Station, TX 77843
b Dep. of Agricultural and Biological Engineering, Clemson Univ., Clemson, SC 29634-0357, and Dep. of Environmental Toxicology and the Clemson Institute of Environmental Toxicology, Clemson Univ., Pendleton, SC 29670

* Corresponding author (Don.Zhu{at}po.state.ct.us)

Received for publication April 22, 2002.

   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
Molecular-level sorption behavior of monoaromatic compounds in suspensions of water-dispersable clay components was studied by measuring 2H nuclear magnetic resonance (NMR) spin–spin relaxation times (T2). In general, decreased T2 values indicate stronger solute–sorbent interactions and increased sorption of the solute. A decreasing trend for T2 values in the order benzene > fluorobenzene > toluene (-C6D5 moiety) was observed, which was probably caused by the hydrophobic effect. The T2 values for benzene and the -C6D5 moiety of toluene increased with increasing pH, whereas the trend with pH was much weaker and less consistent for fluorobenzene and the methyl group of toluene. Conversely, no clear relationship was found between T2 values and pH for dichloromethane. These contrasting results cannot be explained by the pH-dependent self-assembly and hydrophobicity of humics. Instead, directed specific forces, including hydrogen bonding, cation–{pi} interactions, and aromatic–aromatic interactions, are proposed between the benzene ring of monoaromatic solutes and soil organic matter (SOM). Substituents of benzene affect these interactions by varying the {pi} electron density. When the soil fraction was treated with NaOH to remove humic and fulvic acids, T2 values for the different monoaromatic solutes were surprisingly lower compared with those for the untreated soil fraction. This result is probably caused by the increased ratio of solutes adsorbed to "hard" or "glassy" SOM components, which leads to less mobile sorbed solute molecules, after removing NaOH-extractable humics that contain more "soft" or "rubbery" SOM components.

Abbreviations: NMR, nuclear magnetic resonance • NOC, nonionic organic chemical • PAH, polycyclic aromatic hydrocarbon • SOM, soil organic matter • T1, spin–lattice relaxation times • T2, spin–spin relaxation times


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
SORPTION IS A KEY FACTOR controlling the fate of nonionic organic chemicals (NOCs) in the environment. Because of the absence of observational data on molecular-level motions of solutes adsorbed to geosorbents in aqueous systems, hypothesized sorption mechanisms have been advanced based on indirect macroscopic observations (e.g., laboratory sorption isotherms). Previous studies have suggested that sorption of NOCs is a partitioning process between the aqueous phase and the organic phase of geosorbents, which can be quantified by an organic carbon–normalized linear sorption model (Chiou et al., 1979; Karickhoff et al., 1981). To account for later observations of nonlinear sorption, dual-mode sorption models have been proposed (Huang et al., 1997; Xing and Pignatello, 1997). In these models, soil organic matter (SOM) is considered to be heterogeneous and consist of "hard" and "soft" or "glassy" and "rubbery" components that exhibit different behaviors in sorption processes (Young and Weber, 1995). In general, sorption nonlinearity is thought to result primarily from "hard" or "glassy" SOM. The molecular-level mechanisms of solute–sorbent interactions in these macroscopic, phenomenological models, however, are still not clear.

It has been thought that sorption of NOCs is controlled principally by the hydrophobic effect, a combination of relatively small enthalpy effects (e.g., weak van der Waals forces resulting from dipole–dipole, induced dipole–dipole, and induced dipole–induced dipole interactions) and a substantial entropy gradient that drives the organic molecules out of the aqueous phase (Tanford, 1980; Voice and Weber, 1983). Previous research has shown, however, that aromatic chemicals can interact with polar and charged species in aqueous solution through bonding forces with energies larger than those typically exhibited by van der Waals interactions. For example, it has been reported that benzene interacts with water molecules through hydrogen bonding to form 1:1 clusters with a binding energy of about 8 kJ/mol (Suzuki et al., 1992). Spectroscopic data in that study also showed the water molecule to be located above the benzene ring with both hydrogen atoms pointing toward the {pi} electrons of benzene. Recent spectroscopic and modeling studies have shown NH–{pi} hydrogen bonding between aromatic compounds and amine, with interaction intensities increasing with {pi} electron densities (Munoz et al., 2001; Mons et al., 2002). Monoaromatic compounds also have been reported to interact with base cations or charged ammonium or alkylammonia moieties through relatively strong cation–{pi} interactions, which result from electrostatic attractions between the permanent quadrupole of a benzene ring and the cations (Kumpf and Dougherty, 1993; Ma and Dougherty, 1997; Gokel et al., 2001). Soil organic matter is known to have many functional groups (e.g., carboxyl, phenol, amine) that can interact with aromatic compounds through hydrogen bonding or cation–{pi} interactions. Additionally, it has been known for decades that charge-transfer complexes form between {pi} donors and {pi} acceptors through relatively strong aromatic–aromatic interactions (20–30 kJ/mol) (Foster, 1969). Soil organic matter contains subunits of aromatic rings with high electron acceptability (e.g., quinone like structures) that may function as {pi} acceptors under certain conditions (Melcer et al., 1989; J.J. Pignatello, personal communication, 2002). Support for specific interactions with SOM can be found from previous field studies that have observed polycyclic aromatic hydrocarbons (PAHs) to have higher carbon-normalized distribution coefficients (Koc) than predicted by linear partition models (McGroddy and Farrington, 1995; Jonker and Smedes, 2000). Positive correlations between PAH sorption and the aromatic content of sorbents have also been reported (Gauthier et al., 1987; Chin et al., 1997). A more recent study observed high energies (>110 kJ/mol) for PAH sorption to sediments containing a large content of coal and wood-derived constituents (Ghosh et al., 2001). Although most results have been observed only from macroscopic studies (i.e., heterogeneous sorption behaviors of SOM; Accardi-Dey and Gschwend, 2002), they may indicate that relatively strong molecular-level binding forces exist in the sorption of aromatic compounds.

Alternative methods to using sorption isotherms for studying NOC sorption are limited. Fourier transform infrared (FTIR) spectroscopy has been used to study the sorption of certain organic chemicals to humic materials and soils (Martinneto et al., 1994; Landgraf et al., 1998; Suetsugu et al., 2001). However, observing sorbate–sorbent interactions through wavelength shifts of existing bands or appearance of new absorbance bands sometimes can be ambiguous due to high background interferences (e.g., water absorbance). Compared with FTIR, nuclear magnetic resonance (NMR) spectroscopy is able to differentiate signals of interest from the others by probing isotopically enriched chemicals. High-frequency solid-state 13C NMR with magic angle spinning (MAS) has been used to characterize sorption of organic chemicals to dry humics or soils. For example, chemical shifts ({delta}) of 13C-labeled chemicals (e.g., acetone, trichloroethylene [TCE], carbon tetrachloride) have been used to study their sorption on dry minerals, humic acids, and soils (Jurkiewicz and Maciel, 1995).

Deuterium NMR has the advantage of being sensitive to relatively weak solute interactions, and is well suited for NMR relaxation studies because deuterium relaxation is dominated by the quadrupole relaxation mechanism (Smith, 1983). Recently, solution-phase, noncovalent interactions between perdeuterated monoaromatic compounds (e.g., phenol, pyridine, benzene) and natural humic acids have been characterized through measuring spin–lattice relaxation times (T1) (Nanny and Maza, 2001). Spin–spin relaxation times (T2) or the equivalent line-broadening has also been used to characterize sorption of d5–fluorobenzene to organic materials such as surfactant micelles and humic acids (Herbert and Bertsch, 1997). However, no NMR techniques including those mentioned above have been applied to study sorption of NOCs to geosorbents that are suspended in aqueous solution. This may be due to the difficulty of obtaining high-resolution NMR spectroscopy for aqueous soil suspensions.

In this study, high-resolution 2H NMR spectroscopic data was obtained for several monoaromatic compounds and dichloromethane in aqueous suspensions of water-dispersable clay components from a soil. Based on T2 measurements of the different perdeuterated solutes, we conclude that in addition to the hydrophobic effect, directed specific interactions with SOM may be important for the sorption of electron-rich aromatic compounds.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 REFERENCES
 
Chemicals
Perdeuterated compounds including d6–benzene (Cambridge Isotope Laboratories, Andover, MA), d5–fluorobenzene (C/D/N Isotopes, Pointe Claire, QC, Canada), d8–toluene (C/D/N Isotopes), d2–dichloromethane (Aldrich Chemical, Gillingham, UK), and d2–water (Aldrich Chemical) were used as received. All aqueous samples were prepared with doubly distilled water.

Soils
The soil used in this study (collected from Montgomery County, southeastern Texas) is classified as the Tuscumbia series (fine, mixed, nonacid, thermic Vertic Epiaquept), consisting of nearly level, deep, poorly drained, clayey soils. To prepare a water-dispersible soil fraction, 100 g soil was mixed with 10 L water (1:100 soil to water ratio) and shaken on an orbital shaker at 200 rpm for 1 h. After a settling period of 4 h, the fraction with a size of <2 µm (obtained from Stoke's law) was siphoned out from the top 5 cm suspension. The obtained soil fraction was air-dried, and X-ray diffraction (XRD) showed that nearly all of the minerals present were clays (principally smectite). The soil fraction was divided into three separate fractions, one of which was treated with 0.5 M NaOH to extract humic and fulvic acids (Hayes, 1985), and a second which was treated with H2O2 (30%; Fisher Scientific, Pittsburgh, PA) to remove most of the indigenous organic matter (Kunze and Dixon, 1986). All treated and untreated soil fractions were saturated with Na+ by washing four times with 1.0 M NaCl, four times with 0.1 M NaCl, and one time with 0.01 M NaCl. All soil fractions were then freeze-dried and stored in a desiccator until used. Sodium ion saturation was used to increase the suspendibility and hence the homogeneity of the soil suspensions for later NMR analysis. Soil fractions without treatment and with NaOH treatment had measured total organic carbon (TOC) values of 1.41 and 1.32%, respectively, whereas the TOC for the soil fraction receiving H2O2 treatment was below the detection limit.

Sample Preparation
For aqueous solution samples, 1 µL d8–toluene (0.05% v/v) or 2 µL (0.1% v/v) other perdeuterated solute was added to 2 mL 0.01 M NaCl with pre-adjusted pH in a 2-mL glass vial (Teflon-backed screw cap, Fisher). For soil suspension samples, 200 mg soil was added to 40 g 0.01 M NaCl background solution (1:200 soil to solution ratio). After pH adjustment with 0.1 M NaOH or HCl, 1 µL d8–toluene or 2 µL other perdeuterated solute was added to 2 mL soil suspension. All samples were kept in the dark and shaken at ambient temperature for 5 d prior to NMR analysis to allow for sorption equilibration (as verified by independent batch sorption experiments, data not shown).

Nuclear Magnetic Resonance Experiments
Deuterium NMR spectra were recorded at ambient temperature (21 ± 0.5°C) with a Varian (Palo Alto, CA) 400 spectrometer operating at 61.35 MHz, an acquisition time of 2.73 s, and a spectra width of 3000.3 Hz with a 5-mm switchable probe. This high-frequency spectrometer was used to give good shimming, as well as a high signal-to-noise ratio. After locking and primary shimming on a reference sample containing pure d2–water or suspension of d2–water and soil (1:200 ratio) for solution and soil samples, respectively, the spectrometer was run unlocked after fine shimming based on the shape of the peak of interest. The inversion–recovery and the Carr–Purcell Meiboom–Gill (CPMG) pulse sequences were used for T1 and T2 measurements, respectively. A recycle delay of at least 5 T1, and transients of at least 10 were used. Note that all soil factions were well-suspended, with no observable settling after any NMR run (about 5 min after shimming). Spectra were processed with an exponential multiplication corresponding to line broadening of 2 Hz. Chemical shifts were internally referenced to the natural abundance of deuterated water for soil samples (<0.03 ppm variation for the reference water at different pH values, checked with an external reference of deuterated chloroform in water). Values for T1 and T2 were calculated by exponential regressions and reported with instrument-recorded errors.


   RESULTS AND DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
REFERENCES
 
Considerations for Monitoring Sorption Processes via Nuclear Magnetic Resonance Relaxation
Relative properties of the different solutes are summarized in Table 1. No detectable chemical shift (i.e., {delta} > 0.03 ppm) was observed for any deuterated probe in the different soil suspensions. High-resolution NMR spectra were obtained for all probes (representative spectra shown in Fig. 1) . In systems where the deuterium nucleus is rapidly exchanging between two environments of different correlation times, which can be caused by sorption, the observed T2 in the extreme narrowing case can be expressed as (Zens et al., 1976):

where {omega} is the resonance frequency (61.35 MHz), e2Qq/h is the quadrupolar coupling constant, {tau}f and {tau}s are the correlation times for free and adsorbed solutes, respectively, and {chi} is the mole fraction for adsorbed solutes. In general, a decreased T2 value indicates stronger and/or increased sorption of the solute (i.e., higher {tau}s and {chi}). Thus, the sorption behavior of a solute can be related to its observed T2 values.


View this table:
[in this window]
[in a new window]
  		Table 1. Water solubilities, octanol–water partition coefficients (Kow), and dipole moments of the different solutes.

 


View larger version (7K):
[in this window]
[in a new window]
  		Fig. 1. The 2H nuclear magnetic resonance (NMR) spectra of different aromatic compounds and deuterated water at natural abundance in aqueous soil suspensions at pH 7. (a) d6–benzene (downfield-shifted). (b) d5–fluorobenzene (two peaks resolved for {alpha}, the most downfield-shifted, and ß). (c) d8–toluene (one peak resolved for -C6D5 as the most downfield-shifted; -CD3 as the most upfield-shifted).

 
The T2 values of probe molecules are weighted averages of molecules in the adsorbed and dissolved states. Variations in solution chemistry (3 <= pH <= 11; ionic strength < 0.1 M) had no measurable effect on the relaxation of all the probes tested in this study in aqueous solution (as indicated by negligible variation in T1, data not shown). These results also indicated that the potential effect of changing solution viscosity, which affects deuterium relaxation via molecular correlation (Glasel, 1969), was negligible in this study. Additionally, adsorbed molecules are much less mobile and have significantly higher correlation times compared with molecules dissolved in solution (i.e., {tau}s/{tau}f > 104) (Kriz, 2000). Thus, even at very low sorption densities (e.g., {chi} = 0.01), T2 is very sensitive to {tau}s and is principally determined by the terms in the last set of parentheses in Eq. [1] that characterize the contribution from sorption. However, it is important to note that the magnitude of T2 can be significantly decreased by paramagnetic-induced relaxation. No effort was made in the present study to eliminate this effect because of its impracticality (e.g., removal of structured iron in clay). However, since all relaxation times were measured with the same amount of soil in suspension, the same paramagnetic effect would have been present for all samples, thereby resulting in a constant baseline effect.

It is also important to note that relaxation times are dependent on solute structures. This can be seen by the T1 values for the different deuterated solutes in aqueous solution (Table 2). In general, a lower T1 value for 2H NMR indicates a slower molecular rotation characterized by a higher correlation time ({tau}c) (Abragam, 1961). Alkyl deuterium has higher T1 values (faster molecular rotation) than aromatic deuterium because the latter is more restrained in molecular mobility (i.e., benzene deuterium is inhibited from moving out of the planar benzene ring). This can be further demonstrated by the lower T1 value of the ß deuterium of fluorobenzene compared with that of the {alpha} and {gamma} deuterium. This is because the {alpha} and {gamma} carbons have lower electron densities (less aromaticity) than the ß due to the substituent effect of fluorine. Interestingly, the order of benzene > fluorobenzene > toluene is observed for T1. This is because the molecular rotation is also affected by steric effects (i.e., solute size increases in the order toluene > fluorobenzene > benzene).


View this table:
[in this window]
[in a new window]
  		Table 2. Spin–lattice relaxation times (T1) and associated errors (in parentheses) for d6–benzene, d5–fluorobenzene ({alpha}, ß, and {gamma} deuterium), d8–toluene ({alpha}, ß, deuterium, and methyl group), and d2–dichloromethane (CD2Cl2) in 0.01 M NaCl solutions (pH = 7.0).

 
The present NMR spectroscopic method can also be used to a certain degree to monitor sorption linearity. For example, T2 values for d6–benzene at initial concentrations ranging from 5.6 mM (0.05% v/v) to 22.5 mM (0.2% v/v) for the untreated soil fraction and for the one treated with NaOH are shown in Table 3. It can be seen that the sorption of benzene under these conditions is linear because T2 is nearly constant (i.e., constant {chi}).


View this table:
[in this window]
[in a new window]
  		Table 3. Spin–spin relaxation time (T2) (seconds) and associated errors (in parentheses) for d6–benzene in suspensions of soil fractions as a function of added benzene concentration (1:200 soil to solution ratio, 0.01 M NaCl, pH = 7.0).

 
The T2 values for d6–benzene, the benzene ring (-C6D5) of d8–toluene (only one peak identified), and the {alpha} deuterium of d5–fluorobenzene (ß deuterium also identified) for the untreated soil fraction are shown in Fig. 2 . The trend of benzene > fluorobenzene > toluene can be observed for T2, indicating that toluene exhibits the strongest sorption. Dipole–dipole interactions can be ruled out as an explanation for this observation because the observed T2 trend is not consistent with the order of solute dipole moments (e.g., fluorobenzene > toluene > benzene) (Table 1). Instead, this result can be caused by (i) increased toluene sorption (i.e., higher {chi} in Eq. [1]) due to the hydrophobic effect, consistent with water solubilities and octanol–water partition coefficients (Table 1), or (ii) slowed toluene molecular rotation (i.e., higher {tau}s in Eq. [1]) caused by steric effects (toluene > fluorobenzene > benzene). We postulate that the hydrophobic effect is more likely to be the correct interpretation, because the steric differences among these compounds are relatively small due to their similar structures and molecular sizes (i.e., close T1 values in solution, Table 2). In addition, the adsorbed molecules are restrained from free molecular rotation, which would further decrease the importance of steric effects.



View larger version (24K):
[in this window]
[in a new window]
  		Fig. 2. Spin–spin relaxation time (T2) values for d6–benzene ({diamond}), the benzene ring (-C6D5) of d8–toluene ({square}), and the {alpha} deuterium of d5–fluorobenzene ({triangleup}) in suspensions of the untreated soil fraction as a function of pH (1:200 soil to solution ratio, 0.01 M NaCl).

 
Effects of pH on Monoaromatic Compound Sorption
Figure 3 shows T2 values for d6–benzene and -C6D5 of d8–toluene for the untreated soil fraction and for the soil fraction treated with NaOH. Figure 4 shows T2 values for the {alpha} and ß deuterium of d5–fluorobenzene for the same two soil fractions, and Fig. 5 shows them for the methyl group (-CD3) of d8–toluene and d2–dichloromethane. It can be seen that T2 values of d6–benzene and -C6D5 of d8–toluene increase with increasing pH, indicating stronger solute–sorbent interactions at low pH values, whereas much weaker and less consistent trends with pH are evident for d5–fluorobenzene and -CD3 of d8–toluene. For d2–dichloromethane, however, no clear relationship between T2 and pH was observed. Based on line-broadening of d5–fluorobenzene, a previous study suggested that self-assembly of humic acids at low pH caused decreased mobility of the adsorbed solute (Herbert and Bertsch, 1997). However, because of the relative stiffness of the solid matrix, SOM may not be able to rearrange its conformation as readily as humic materials dissolved in aqueous solution. A recent study by Nanny and Maza (2001) also observed decreased mobility of solutes in humic acid solution at low pH, based on T1 measurements of d6–benzene. The investigators proposed that protonation of humic acids at low pH increased their hydrophobicity and hence their hydrophobic interactions with the solute. This hypothesis, however, is not likely to be the major reason for the T2–pH relationship observed in the present study, because otherwise fluorobenzene T2 values would be more sensitive to pH variations than benzene due to its more hydrophobic characteristics. Additionally, the two hypotheses above cannot explain why the T2–pH relationship observed here is dependent on solute structures. For example, T2 values for the -CD3 moiety of toluene, the ß deuterium of fluorobenzene, and dichloromethane exhibit trends different from those for benzene and the -C6D5 moiety of toluene.



View larger version (25K):
[in this window]
[in a new window]
  		Fig. 3. Spin–spin relaxation time (T2) values for d6–benzene and -C6D5 of d8–toluene in suspensions of the untreated soil fraction and the soil fraction treated with NaOH as a function of pH (1:200 soil to solution ratio, 0.01 M NaCl). {diamond} d6–benzene for untreated soil; {square} -C6D5 of d8–toluene for untreated soil; {diamondsuit} d6–benzene for treated soil; {blacksquare} -C6D5 of d8–toluene for treated soil.

 


View larger version (24K):
[in this window]
[in a new window]
  		Fig. 4. Spin–spin relaxation time (T2) values for the {alpha} and ß deuterium of d5–fluorobenzene in suspensions of the untreated soil fraction and the soil fraction treated with NaOH as a function of pH (1:200 soil to solution ratio, 0.01 M NaCl). {diamond} {alpha} d5–fluorobenzene for untreated soil; {square} ß d5–fluorobenzene for untreated soil; {diamondsuit} {alpha} d5–fluorobenzene for treated soil; {blacksquare} ß d5–fluorobenzene for treated soil.

 


View larger version (25K):
[in this window]
[in a new window]
  		Fig. 5. Spin–spin relaxation time (T2) values for -CD3 of d8–toluene and d2–dichloromethane in suspensions of the untreated soil fraction and the soil fraction treated with NaOH as a function of pH (1:200 soil to solution ratio, 0.01 M NaCl). {diamond} -CD3 of d8–toluene for untreated soil; {triangleup} d2–dichloromethane for untreated soil; {diamondsuit} -CD3 of d8–toluene for treated soil.

 
Another possible explanation for the results observed here is that pH affects T2 values by varying the concentration of dissolved organic carbon (DOC) released from the soil fractions to solution. For example, it is well known that increases in pH will increase the dissolution of SOM (You et al., 1999), which may decrease solute sorption to the solid phase (i.e., higher T2). However, this phenomenon is not likely to be the correct explanation, either, because it is apparent that T2 values for fluorobenzene with the NaOH-treated soil fraction are even more pH dependent than those with the untreated soil fraction (Fig. 4). Because the NaOH-treated fraction was subjected to pH conditions much higher than pH 9, its remaining SOM would presumably be more resistant to dissolution. For the other solutes, the observed T2–pH relationship was also generally stronger for the NaOH-treated soil fraction. Another possibility is that pH affects solute T2 values via varying physical–chemical properties of the clay surface. For example, a recent study observed pH effects on the adsorption of dinitrophenol herbicides to montmorillonites by modifying the charge density, as well as the size of adsorption domains (Sheng et al., 2002). Changes in the clay surface properties with pH would also be expected to result in different T2–pH trends for the different solutes. However, the hypothesis is probably not the correct explanation for our observed trends either because the clay surface is only a minor contributor to the overall sorption of monoaromatic compounds, as was confirmed by experiments conducted with an H2O2–treated soil fraction (see below).

We propose that the apparent pH-dependent sorption observed for these monoaromatic compounds is caused by directed molecular-level solute–sorbent interactions, including hydrogen bonding, cation–{pi} interactions, and aromatic–aromatic interactions, that are either acting alone or, more likely, together in some cooperative fashion. Although no observational data are available to date, we hypothesize that hydrogen atoms of functional groups (e.g., carboxyl, phenol, ammonia) in SOM interact with benzene rings of aromatic solutes, especially those with high {pi}-electron densities (e.g., alkyl-substituted benzene, PAHs), through hydrogen bonding. Considering that most functional groups (e.g., protonated carboxyl) are good candidates for hydrogen bonding, this hypothesis appears plausible.

The potential importance of these and other so-called cation–{pi} interactions has been recognized for some time in the biological sciences (Dougherty, 1996; Ma and Dougherty, 1997; Gokel et al., 2001). For example, cation–{pi} interactions between amino acid side chains function as a mechanism of intermolecular recognition at the protein–protein interface in aqueous solution (Pletneva et al., 2001). The ionizable moities in SOM include amino and ammonia groups (>50% of identified nitrogen in humics) (Schnitzer, 1985). Here, we hypothesize that positively charged nitrogen groups interact with aromatic solutes through cation–{pi} interactions (i.e., charged amino groups face the benzene ring). Despite the low content of nitrogen in SOM (generally less than a few percent) (Schnitzer, 1985), cation–{pi} interactions invoked by the nitrogen groups can still be important to sorption of aromatic solutes because of the relatively high interaction intensities (>80 kJ/mol; Ma and Dougherty, 1997).

The role of aromatic–aromatic interactions in the sorption of aromatic pollutants to SOM is currently under investigation by another research group (J.J. Pignatello, personal communication, 2002). Recent results from their work indicate that strong face-to-face aromatic–aromatic interactions occur between the {pi} donor (phenanthrene) and the {pi} acceptor (diprotonated o-phenanthroline) in aqueous solution (J.J. Pignatello, personal communication, 2002). Although the monoaromatic chemicals tested in this study are weak electron donors compared with phenanthrene, aromatic–aromatic interactions with {pi} acceptors in SOM (e.g., quinone like structures) (Melcer et al., 1989; J.J. Pignatello, personal communication, 2002) that have energies higher than van der Waals interactions are likely to occur.

Directed specific forces including cation–{pi} interactions and aromatic–aromatic interactions are very strong in organic solvents (Ma and Dougherty, 1997; Foster, 1969). In aqueous solution, however, water can often interfere with these interactions through competitive reactions (e.g., hydrogen bonding with the functional groups and cation hydration), leading one to mistakenly believe that these interactions are no longer important for NOC sorption from aqueous solution. In fact, however, relatively strong cation–{pi} interactions and aromatic–aromatic interactions have been observed in aqueous solution (Ma and Dougherty, 1997; Gokel et al., 2001; J.J. Pignatello, personal communication, 2002). Additionally, a previous study using molecular dynamic simulations has shown that benzene caged in a clathrate hydrate in aqueous solution forms stronger hydrogen bonds than in pure water because the rigid host framework facilitates orientation of the benzene molecule (Fujii et al., 1998). Similarly, sorption domains that trap solute molecules may help keep them anchored and oriented to favorable directed specific interactions.

Variations in pH can dramatically affect hydrogen bonding, cation–{pi} interactions, and aromatic–aromatic interactions. For example, low pH values favor hydrogen bonding as well as cation–{pi} interactions because of the speciation of the functional groups (i.e., protonated carboxyl groups and positively charged amino groups at low pH values). Low pH values may also favor aromatic–aromatic interactions (J.J. Pignatello, personal communication, 2002). The {pi} acceptors in SOM have high electron acceptability at low pH values because the attached carboxyl groups are protonated (i.e., -COOH draws electrons), whereas they become weaker or non-{pi} acceptors at high pH values because these groups are deprotonated (i.e., -COO- donors electrons). Also, the {pi}-electron densities of aromatic solutes affect the intensities of hydrogen bonding, cation–{pi} interactions, and aromatic–aromatic interactions, because they all function via electrostatic attraction. Therefore, the higher the {pi} electron density of an aromatic solute, the stronger the interaction. Because fluorobenzene has a lower {pi} electron density due to its substituent, its T2 value is less sensitive to pH variations compared with d6–benzene and the -C6D5 moiety of d8–toluene (Fig. 3). Conversely, dichloromethane and the -CD3 moiety of toluene cannot interact with SOM through those specific interactions. Because only dipole interactions are involved, pH has less of an effect on the T2 values for these species (Fig. 5). It should be noted that T2 of the -CD3 moiety of toluene is more dependent on its own relaxation than on that of the benzene ring due to the freely rotating C–C {sigma} bond. However, T2 of -CD3 is less sensitive to pH variation because the group apparently does not interact much with SOM.

Effects of Different Soil Organic Matter Domains on Monoaromatic Compound Sorption
In Fig. 3 through 5, it can be seen that T2 values are lower for the soil fraction treated with NaOH than the untreated soil fraction for all tested solutes. To further test the effect of SOM on sorption, T2 of d6–benzene for a soil fraction treated with H2O2 (1:200 solid to solution ratio, 0.01 M NaCl, pH = 7.0) was also measured. The value (T2 = 0.1499 ± 0.0148 s) was higher than that for the untreated soil fraction, indicating a decreased sorption capacity after removal of SOM. Therefore, since partial removal of SOM by NaOH extraction would not have led to an overall higher sorption capacity (higher {chi} in Eq. [1]), we conclude that decreased molecular mobility of adsorbed solutes must be responsible for the decrease in observed T2 values for the soil fraction treated with NaOH.

The conclusion above is consistent with proposed dual-mode sorption models (Huang et al., 1997; Xing et al., 1996; Xing and Pignatello, 1997). In these models, SOM is thought to have "hard" and "soft" or "glassy" and "rubbery" domains (Young and Weber, 1995). The "soft" or "rubbery" component behaves similarly to a partition solvent, whereas the "hard" or "glassy" component, due to its rigidity, possesses nanoscale pores that can serve as adsorption sites. Humin contains highly condensed, relatively polar heteroaromatic or methylene backbones (Preston and Newman, 1992; Almendros et al., 1998; Weber et al., 2000), which can be considered "harder" or more "glassy" than NaOH-extractable humic and fulvic acids that have relatively low molecular weights (Xing, 2001). Because clayey soils tend to be rich in aliphatic carbon (Theng et al., 1986; Preston, 1996), it can be inferred that the SOM of the soil fraction tested in this study is rich in condensed, polar aliphatic backbones (e.g., ester-like structure) because of its high humin content. Also, the ratio of solutes adsorbed to the "hard" or "glassy" SOM would be higher for the soil fraction treated with NaOH compared with the untreated soil fraction. Once trapped in the rigid pores of the "hard" or "glassy" SOM, solute molecules would be less mobile than those partitioned to the "soft" or "rubbery" SOM, resulting in lower solute T2 values for the NaOH-treated soil fraction even though it may exhibit less overall solute sorption. The difference in T2 values for the untreated soil fraction and the one treated with NaOH also indicates that NaOH-extractable (e.g., humic and fulvic acids) and nonextractable SOM (humin) represent different sorption domains.

The positive trend between T2 and pH that was observed for the untreated soil fraction is also observed for the soil fraction treated with NaOH (Fig. 3–5). For fluorobenzene, this relationship becomes more apparent for the treated soil. There are several possible reasons for this. First, the humin residue after NaOH extraction still contains a high content of functional groups that can participate in directed specific interactions. It has been reported that the content of amino and ammonia moieties in humin is comparable with that in humic and fulvic acids (Schnitzer, 1985). Although aliphatic chains may be the dominant structures in humin, it probably contains fairly significant amounts of carboxylic and aromatic groups (Preston et al., 1989; Grasset et al., 2002). Second, during the NaOH treatment process, free functional groups would have been produced due to the hydrolysis of SOM subunits catalyzed by the strong base condition (i.e., carboxyl and amino groups from hydrolysis of peptides). This directly enhances specific interactions, as well as the dipole-involved interactions with relatively strong intensities. For example, fluorobenzene can interact with charged amino groups at low pH via charge–dipole interactions, which would typically be stronger than a regular dipole–dipole interaction (Israelachvili, 1985). This may explain why T2 values for {alpha} and ß deuterium of fluorobenzene were very low at low pH for the NaOH-treated soil fraction. Finally, the rigid pores in "hard" or "glassy" components facilitate orientation of the adsorbed solute molecules to favor directed specific interactions. Under these conditions, specific interactions may become more sensitive to further perturbations resulting from pH variations.

CONCLUSIONS
Sorption of aromatic compounds is driven by the hydrophobic effect, directed specific interactions, and other relatively weak interactions (e.g., van der Waals attractive forces). The roles and relative importance of these driving forces may vary with respect to different sorption stages and/or physical–chemical properties of the sorption system. After aromatic molecules are driven from aqueous solution via the hydrophobic effect, molecular-level interactions relocate and stabilize the solute molecules locally in the sorption domain. During this stage, conformations of the sorption domain can be modified by specific solute–sorbent interactions. For example, a positively charged amino group on an alkyl chain may approach an adsorbed PAH molecule to participate in a cation–{pi} interaction. Similarly, an adsorbed PAH molecule may reorient an aromatic SOM subunit via an aromatic–aromatic interaction. Directed specific interactions may be very important when considering long-term sorption processes such as "aging" (Young and Weber, 1995). A previous study has suggested that humic acids may have membrane-like structures, with the hydrophobic components arranged in the interior of a pseudo-micelle (Wershaw, 1993). Thus, directed specific interactions may become more important after some time when solute molecules access interior SOM domains that may be more like an organic solvent, which would then enhance these interactions.


   ACKNOWLEDGMENTS
 
We thank Ms. LaiMan Lee for preparation of the water-dispersable soil and the soil classification test, and the U.S. Department of Energy for financial support.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
D. Zhu, current address: Connecticut Agricultural Experiment Station, 123 Huntington St., New Haven, CT 06504.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 




This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
D. Zhu, B. E. Herbert, M. A. Schlautman, E. R. Carraway, and J. Hur
Cation-{pi} Bonding: A New Perspective on the Sorption of Polycyclic Aromatic Hydrocarbons to Mineral Surfaces
J. Environ. Qual., July 1, 2004; 33(4): 1322 - 1330.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
D. Zhu, B. E. Herbert, M. A. Schlautman, and E. R. Carraway
Characterization of Cation-{pi} Interactions in Aqueous Solution Using Deuterium Nuclear Magnetic Resonance Spectroscopy
J. Environ. Qual., January 1, 2004; 33(1): 276 - 284.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (4)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zhu, D.
Right arrow Articles by Schlautman, M. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zhu, D.
Right arrow Articles by Schlautman, M. A.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Zhu, D.
Right arrow Articles by Schlautman, M. A.
Related Collections
Right arrow Isotopes
Right arrow Organic Compounds
Right arrow Soil Pollution
Right arrow Water Pollution

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome