Thermodynamics of Peat-, Plant-, and Soil-Derived Humic Acid

Humic acids (HAs) form coatings on clays and minerals that can play an important role in nutrient and contaminant migration in soil and water. Humic a...
0 downloads 0 Views 109KB Size
Download PDF
Environ. Sci. Technol. 2004, 38, 3338-3342

Thermodynamics of Peat-, Plant-, and Soil-Derived Humic Acid Sorption on Kaolinite ELHAM A. GHABBOUR,* GEOFFREY DAVIES, MELISSA E. GOODWILLIE, KELLY O’DONAUGHY, AND TAMMY L. SMITH Department of Chemistry and Chemical Biology, Northeastern University, Boston, Massachusetts 02115-5000

Humic acids (HAs) form coatings on clays and minerals that can play an important role in nutrient and contaminant migration in soil and water. Humic acid-clay mineral interactions are known to be affected by pH and ionic strength, but little attention has been paid to the effects of temperature. In this paper we report the stoichiometry and thermodynamics of interactions of aqueous HAs (isolated from two peats, two soils and a marine alga with a method that removes lipids) with kaolinite clay, Al2Si2O5(OH)4, at seven temperatures from 5.0 to 35.0 °C in 0.05 M NaCl at pH 3.5. All the sorption isotherms exhibit consecutive steps ascribed to HA monolayer and bilayer formation, respectively. Site capacity comparisons suggest different HA molecular conformations on kaolinite. Linearly correlated enthalpy and entropy changes for HA sorption point to the importance of hydration and dehydration in the sorption mechanism.

Introduction Habitable land is thermally balanced with the oceans by lakes, rivers, vegetation, and soils. Soil organic matter (SOM) retains water, partly because it contains humic acids (HAs), the highly functionalized metabolites of animal and plant biopolymers (1). HAs also play key roles in metal binding, organic solute sorption, and biomineralization (2, 3). Preserving and remediating land and water resources depends in large part on knowing the HAs’ origins, structures, and properties (1). HAs typically are anchored in SOM by sorption on clays and minerals (4, 5), which decreases HA solubility, improves soil texture, and lessens soil erosion. Humic acid-clay mineral interactions are affected by pH and ionic strength (6-10), but little attention has been paid to the effects of temperature. Isotherm data for nucleic acid constituent (NAC) solute sorption (11, 12) and metal binding (2, 13) by lipid-depleted, solid HAs fit the Langmuir model and indicate consecutive NAC sorption and metal binding steps labeled i ) A, B, and C. For both NAC solute sorption and metal binding we found linearly correlated enthalpy-entropy pairs [∆Hi, ∆Si], indicating that HAs behave as free energy buffers in their interactions with NACs and metals (11-13). HA sorption on a clay or mineral also is a heterogeneous process, and the thermodynamic parameters might also be correlated. In this paper we report the stoichiometry and thermodynamics of * Corresponding author phone: (617) 373-7988; fax: (617)-3738795; e-mail: [email protected]. 3338

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 12, 2004

sorption of five lipid-depleted aqueous peat-, plant-, and soil-derived HAs on kaolinite, Al2Si2O5(OH)4, at pH 3.5 and ionic strength 0.05 M (NaCl). The measurements were made at seven temperatures in the range 5.0-35.0 °C. The objectives were (1) to investigate the possibility of multilayer HA coatings on kaolinite, as observed on minerals (14) and metal oxides (15), and (2) to see if the enthalpy and entropy changes for HA sorption on kaolinite are linearly correlated.

Experimental Section Materials. Lipids can block sorption sites for hydrocarbons and other solutes in HAs, and their removal with benzene/ methanol prior to HA isolation dramatically affects HA sorption behavior (6). We included this preextraction step (16) in the isolation of solid HAs used in previous studies of NAC sorption (11, 12) and metal binding (2, 13). HA products isolated from the same source after preextraction with 2:1 benzene/methanol (16) or sequentially with diethyl ether, acetone, ethanol, dioxane, and water (17) are closely similar. Since HAs originate from different sources, we compare the properties of large, representative HA samples isolated from composts, plants, peats, and soils. Two rich HA sources are an Irish peat from the Turf Board Co. (Cork, Ireland) that gives humic acid IHA and a bog soil from the White Mountain National Forest (Rumney, NH) that gives NHA (2). Humic acids GHA and NYHA were isolated with the same mild protocol (2, 12, 16) from a peat supplied by Staatsbad Pyrmont AG (Bad Pyrmont, Germany) and an organic farm soil in Pine Island, NY, respectively. This protocol with additional steps to remove uronic acids (18, 19) was used to isolate PHA from the abundant free-living marine alga Pilayella littoralis in ocean water north of Boston, MA, as previously described (18). HA solutions were made by stirring 100.0 mg of each freeze-dried solid HA in 1.00 L of 0.05 M NaOH at 25 °C until the solid was completely dissolved. Each solution was then adjusted to pH 7.0 or 3.5 by addition of concentrated HCl. The stock solutions were diluted at pH 3.5 to give solutions containing 5.0, 10.0, 20.0, 30.0, 40.0, 50.0, 60.0, 70.0, 80.0, and 90.0 mg of HA/L. These solutions were used to plot optical absorbance (at 280 nm) versus HA concentration to calibrate absorbance measurements at 280 nm used to construct sorption isotherms at seven fixed temperatures from 5.0 to 35.0 °C. All other chemicals were analytical reagent grade. Doubly deionized water was used throughout. Kaolinite. A large sample of well-crystallized kaolinite from the Geology Department Reference Collection at Northeastern University was purified as follows (20). The freeze-dried sample was ground in an agate mortar and fractionated dry with geological sieves to give a fraction with average particle size 20 ( 2 µm. The sieved sample was treated with 1.0 M acetate buffer (pH 5) at 70 °C and then with excess 1.0 M H2O2 to remove carbonate, organic matter, and MnO2. After the sample was washed with warm acetate buffer, methanol, and 95% aqueous acetone, iron oxides were removed with 0.3 M trisodium citrate, sodium bicarbonate, and sodium thiosulfate at 80 °C followed by addition of saturated NaCl solution, centrifugation at 2500 rpm, and washing twice with 1 M NaCl solution. The >2 µm kaolinite particles were resuspended in deionized water and dialyzed (membrane cutoff 10 kDa) against deionized water until free of chloride. Finally, the purified kaolinite was freeze-dried before use in HA sorption/desorption experiments and solubility tests. Its surface area was measured in a custombuilt instrument (21) with nitrogen as adsorbate at 77 K. 10.1021/es0352101 CCC: $27.50

 2004 American Chemical Society Published on Web 05/15/2004

FIGURE 1. FTIR spectra of kaolinite and HA-treated kaolinite samples in KBr disks. Sorption Measurements. Sorption of each HA on kaolinite was studied as follows. Separate vessels containing 20.0 mg of solid kaolinite were treated with 20.0 mL of each of the 10 HA solutions listed above. Fixed pH 3.5 was used for most of the work because HA sorption on kaolinite increases with decreasing pH (8). The mixtures were shaken for 48 h at fixed temperature in a thermostated shaker, and the absorbance of the supernatant of each mixture was measured at 280 nm after centrifugation for 1 h at 2000 rpm. The calibration curve for each HA was used to calculate the concentration c (g/L) of the HA in equilibrium with solid kaolinite and the amount of HA sorbed from each mixture (A, g of HA/g of kaolinite). The results were plotted as the isotherm A versus c for each HA at each temperature. Equilibration of each HA with kaolinite occurred within 24 h of mixing in all cases. Isotherms were repeated over a period of several months to ensure complete HA disaggregation (22) and equilibration with kaolinite. Data Analysis. The Langmuir model (11, 23) predicts that data for a single-step sorption process (eq 1) on solid kaolinite will obey eq 2. Here, v is the site capacity (g of HA/g of kaolinite) of kaolinite for the solute and K is the equilibrium constant for the sorption process.

kaolinite(s) + HA(aq) a kaolinite‚HA(s) 1/A ) l/v + 1/Kvc

(K)

(1) (2)

At fixed temperature, eq 2 predicts a linear plot of 1/A versus 1/c with positive intercept 1/v and slope 1/Kv. Separate values of v and K can be calculated. Linear segments in plots of eq 2 indicate successive sorption steps with increasing c (2, 11-13). Kaolinite has a fixed number of sites per gram, so v is independent of temperature (11, 23). Equation 3 describes how equilibrium constant K should vary with absolute temperature T for an equilibrium system. Here, R is the ideal gas constant. Equation 4 predicts a linear plot of log K versus 1/T for reversible sorption of an HA on a fixed kaolinite site. The slope of eq 4 is -∆H/2.303R, and the intercept is ∆S/2.303R.

RT ln K ) T∆S - ∆H

(3)

log K ) -∆H/2.303RT + ∆S/2.303R

(4)

Kaolinite Solubility and HA Desorption. HAs and kaolinite are soluble in aqueous strong bases, which potentially can remove (desorb) HA coatings on clay or mineral surfaces. Removal of each HA from kaolinite by 0.10 M NaOH was checked by absorbance measurements at 280 nm to detect HA removal at the highest experimental temperature (35.0 °C) with continuous sample shaking for 7 days. The supernatants’ pH values were reduced to 3.5 with concentrated HCl, and the supernatants were examined for cloudiness or precipitation of kaolinite on standing overnight.

Results and Discussion Reactants. Kaolinite (Al2Si2O5(OH)4) occurs as clusters of rhombic or hexagonal plates with accessible, hydrophilic surface OH groups and little compositional variation (24). References 2 and 13 give analytical and spectral data for the HAs of this study. Table S1 (in Supporting Information) shows that GHA, IHA, NHA, and NYHA have similar C, H, N contents and low ash contents, but PHA isolated (16, 18, 19) from P. littoralis has lower C and O and higher N and ash contents than the other HAs. Table S1 (in Supporting Information) shows that all five HAs have similar carboxylic acid group contents. However, IHA and PHA are more acidic than the other HAs because of their higher phenolic content. The 1H and 13C NMR spectra of solids IHA, NHA, and PHA are quantitatively similar (2, 25). Percentages of different 13C types indicate that, on average, 9% of the detected carbon is carboxylic. Sorption Behavior. The major effects of sorbing GHA and other HAs on minerals are loss of the characteristic 2500 and 1715-1730 cm-1 FTIR features of HAs (2, 5, 26), a shift to lower frequency and broadening of the doublet that includes the latter feature, and changes in the region below 1050 cm-1 (Figure 1). These effects indicate sorption on minerals through -COO- functional groups (2, 5, 26). Plots of optical absorbance at 280 nm versus HA concentration were linear for solutions of all the HAs to at least 100 mg of HA/L and had nearly the same slope at pH 3.5 and 7.0 for a given HA (Figure S1 in Supporting Information). There was no optical evidence for HA aggregation and no changes of slope or linearity over a period of 18 months. Avena and Wilkinson (22) reported month-long disaggreVOL. 38, NO. 12, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3339

FIGURE 2. Isotherms for sorption of humic acids on kaolinite at ionic strength 0.05 M (NaCl), pH 3.5, and 25 °C. gation of 5.28 g of Irish-peat-derived HA/L stored at pH 3.2 when it was diluted to 30 mg of HA/L at pH 3.6. However, their stock solution was >50 times more concentrated than ours, whose concentration never exceeded 100 mg of HA/L at pH 8.5, where HA disaggregation is rapid (22). Thus, HA disaggregation does not interfere with sequential sorption of HAs on kaolinite described below. The slope of the calibration curve for PHA was the same at pH 3.5 and 7.0 but was smaller than those for GHA, IHA, and NHA (Figure S1 in Supporting Information). NMR data indicate that PHA is more aliphatic than the other HAs (Table S2 in Supporting Information) (25). Figure 2 and Figures S2 and S3 (in Supporting Information) show representative sorption isotherms. Kaolinite sorbs more IHA and PHA per g of kaolinite than GHA and NHA at 25.0 °C (Figure 2). Sorption of NHA (Figure S2 in Supporting Information) and PHA (Figure S3 in Supporting Information) on kaolinite is endothermic, as found for most of the detected sorption steps. The isotherms for IHA (Figure 2), NHA (Figure S2 in Supporting Information), and PHA (Figure S3 in Supporting Information) have evidence for two sorption steps on kaolinite, whereas two steps for NHA and GHA are not easy to see on the vertical scale of Figure 2. This is only evident when eq 2 is plotted and results in two lines with different slopes at low (step A) and higher (step B) values of c at different temperatures, as demonstrated in Figure 3. Site capacities vi and equilibrium constants Ki calculated from the slopes and intercepts of the segmented plots are collected in Table S3 (in Supporting Information). The respective thermodynamic parameters ∆Hi and ∆Si from plots of eq 4 (Figure 4) are included in Table S3 in Supporting Information. Different thermodynamic behavior at high and lower temperatures for a few HA-NAC solute sorption systems has been noted previously as dual temperature behavior (see, for example, ref 12) and is reported for step B of sorption of NYHA on kaolinite (Table S3 in Supporting Information) and step A of GHA sorption (Figure 4). Table 1 gives all the isotherm parameters. Site Capacities. Step A. The surface area of the kaolinite sample was 52 ( 3 m2/g (three determinations (21)). The site capacity data vA in Table S3 in Supporting Information represent the stoichiometry of monolayer formation by a given HA. Except for two systems (IHA and PHA at 5.0 and 10.0 °C), site capacities vA are independent of temperature within experimental precision and average well for a given HA (Table 1 and Table S3 in Supporting Information). This is consistent with a fixed kaolinite surface area for HA monolayer formation. The larger capacity 〈vA〉 ) 0.092 and 0.11 g of HA/g of kaolinite for IHA and PHA, respectively, than for GHA and NHA (Table 1) suggests that IHA and PHA molecules occupy less of the kaolinite surface at saturation. One explanation for this result is that sorbed IHA and PHA 3340

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 12, 2004

FIGURE 3. Plots of eq 2 for steps A and B of sorption of NHA on kaolinite at ionic strength 0.05 M (NaCl) and pH 3.5 at 25.0 and 35.0 °C.

FIGURE 4. Plots of eq 3 for steps A and B of sorption of PHA and step A of sorption of GHA on kaolinite at ionic strength 0.05 M (NaCl) and pH 3.5.

molecules are longer than they are wide. A sausage-shaped surface HA conformation is featured in Wershaw’s membrane-micelle model for HA interactions with minerals through carboxylate groups (16, 26). Detailed studies of the interactions of the HAs of the present work with other clays and minerals (e.g., bauxite) (26) may reveal that GHA and NHA molecules always occupy more of the sorbent surface than IHA and PHA. Evidence for different molar masses or molecular shapes in solution for HAs from different sources also will help to explain the differences reported here. According to X-ray scattering data, the surfaces of humin are dominated by micro- and mesopores (27). We note that 〈vA〉 increases in the order GHA ≈ NHA ≈ NYHA < IHA ≈ PHA, which is roughly the order of increasing phenolic -OH content (Table S1 in Supporting Information).

TABLE 1. Site Capacities and Thermodynamic Parameters for Sorption of HAs on Kaolinite at pH 3.50 and Ionic Strength 0.05 M (NaCl) step A HA GHA NHA NYHA IHAg PHA

〈vA〉a 0.039 0.045 0.051 0.092 0.11

Kb 630 585 740 3100 175

step B

∆Hc -18.3,e 9.5 10.9 15.4 -11.4

6.2f

step C

∆ Sd

〈vB〉a

Kb

∆H c

∆S d

-51,e 34f

0.050 0.067 0.14 0.16 0.12

155 58 17.4 98 144

3.5 11.1 -9.0,e 14.4f -6.5 10.0

22 46 -26,e 54f -13 43

45 50 67 -28

〈vC〉a

0.09

∆Hc

19.1

∆S d

69

∑〈vi〉a 0.089 0.11 0.19 0.34 0.23

a Units are g of HA/g of kaolinite. b Units are L/g of HA at 25.0 °C. c Units are kcal mol-1. d Units are cal deg-1 mol-1. e At low temperature. f At high temperature. g For IHA in step C, K )12.5 L/g of HA at 25.0 °C, with ∆Hc ) 19.1 kcal mol-1 and ∆Sc ) 69 cal deg-1 mol-1 at 25 °C (see Table S3 in Supporting Information).

Steps B and C. Site capacity vB for a given HA may increase slightly with increasing temperature (see, for example, data for IHA sorption in step B, Table S3 in Supporting Information). However, within measurement precision the vB data for a given HA average well over the experimental temperature range to give average 〈vB〉 data in Table 1. The site capacity trends ascribed to HA monolayer formation from 〈vA〉 comparisons also are observed in sorption step B, with 〈vB〉 increasing with solutes in the order GHA ≈ NHA ≈ NYHA < PHA ≈ IHA (Table 1). Use of higher IHA equilibrium concentrations c enabled us to detect sorption of a third IHA layer on kaolinite (Table 1 and Table S3 in Supporting Information). The effect of different HA molecular conformations on the site capacities of monolayer and subsequent HA sorption steps is consistent with the trends in 〈vA〉, 〈vB〉, and the total site capacities ∑vi ) 〈vA〉 + 〈vB〉 + 〈vC〉 (Table 1). The latter increase with HA solutes GHA ≈ NHA < NYHA < IHA ≈ PHA (Table 1). On the basis of the present information, IHA and PHA are proposed to form bilayers on kaolinite. NamjesnikDejanovic and Maurice (14) have observed natural organic matter spheres and bilayers on minerals using atomic force microscopy, and Vinodgopal and co-workers have produced HA bilayers on indium-tin oxide electrophoretically (15). Equilibrium Constants. Equilibrium constants KA for monolayer sorption of the five HAs of this study on kaolinite vary from 175 to 3100 L/g at 25.0 °C and increase in the order PHA < NHA ≈ GHA ≈ NYHA < IHA (Table 1). The equilibrium constants KB at 25.0 °C for HA bilayer formation range from 58 to 155 L/g and increase in the order NYHA < NHA < IHA < GHA ≈ PHA (Table S3 in Supporting Information). The different orders suggest different conformational and/or functional group interactions between different HAs and kaolinite functional groups and between HA molecules themselves. The affinity of IHA for “bilayer” IHA is weak (KC is 12.5 L/g at 25 °C, Table S3 in Supporting Information) and is not observed with the other HAs under the same conditions. Given the possibility that HAs from different sources have different (perhaps temperature-dependent) shapes and functional group distributions, comparisons of sorption equilibrium constants at fixed temperature have limited mechanistic value. Previous work (7, 8) supports our results. Kretzschmar et al. (7) found their HA solutions obeyed the Beer-Lambert law up to 6 mg of HA/L and no evidence for preferential sorption on kaolinite of different 〈Mw〉 HA fractions from their unfractionated HA sample. Measurements with equilibrium HA concentration up to 5.3 mg of HA/L at 25 °C gave isotherms that fit eq 2 with one sorption step ascribed to monolayer formation. The maximum sorbed amount of HA (A, g of HA/g of kaolinite) decreased with increasing pH from 4 to 9 and with increasing ionic strength, and was 0.07 g of HA/g of kaolinite at pH 4 and 25 °C, to be compared with the data in Table 1. It is well-known that pH and ionic strength affect HA molecular configurations and sorption behavior.

Zhou et al. (8) reported that sorption of HA fractions at added concentrations up to 50 mg of HA/L on a range of clays and minerals was most pronounced for HAs with high aromaticity (a soft-coal-derived HA from Aldrich was sorbed to the extent of about 0.022 g of HA/g of kaolinite at ionic strength 1.0 M, 25 °C, and pH 8) and decreased with increasing pH, decreasing ionic strength, and decreasing kaolinite surface area. Although Zhou et al. (8) made no detailed analysis, the isotherms reported by Kretzschmar (7) and Zhou (8) refer to monolayer formation with a Langmuir mechanism. Significantly, Zhou et al. (8) reported that the amounts A of Aldrich HA and a water-derived HA sorbed by kaolinite increased with increasing temperature in artificial seawater at pH 8, while those of water-derived fulvic acid and hydrophilic fractions decreased with increasing temperature in the same medium. Thus, sorption of humic substances from different sources on clays and minerals can be endothermic or exothermic, but aside from the qualitative work of Zhou et al. (8), we have found no thermodynamic parameters for HA-clay and -mineral interactions. Thermodynamic Parameter Correlation. Previous work (11, 12) indicates that 14 aqueous NACs adsorb in consecutive steps with increasing equilibrium solute concentration c on solid compost-derived HA. The enthalpy and entropy changes for each solute and step (including data for two systems that exhibit dual temperature behavior) are correlated on a single line. Site capacity and thermodynamic parameter variations with these 14 solutes were used to develop a “site creation” model for their sorption on compost-derived HA (11). This model points to the need to dehydrate hydrophilic HA and NAC functional groups for sorption to occur. The thermodynamic data for uracil, uridine, and uridine-5′-monophosphate sorption on GHA and GHA saturated with Hg(II) (MGHA) (12) fit a linear correlation, and sorption of uracil, uridine, and uridine-5′-monophosphate on GHA and MGHA is, on average, 4.6 kcal mol-1 more endothermic than on compost-derived HA (12). Our data show that stepwise sorption of aqueous HAs on kaolinite and on monolayer HA is mostly endothermic and results in an entropy increase (Table 1). In the absence of other factors, an entropy decrease is expected for sorption of a dissolved solute on a solid phase. Figure 5 shows that the enthalpy and entropy changes for sorption of the five HAs of this study in all detected steps are linearly correlated. The origin of the linear correlation is the much larger ranges of ∆Hi (-18.3 to +19.1 kcal mol-1) and ∆Si (-51 to +67 cal deg-1 mol-1) than of ∆Gi in eq 5. The slope of Figure 5 is 292 K (23 °C), and the intercept is -2.7 kcal mol-1. Separate plots of ∆HA versus ∆SA and ∆HB versus ∆SB have slopes of 290 K (17 °C) and 298 K (25 °C) and intercepts of -3.3 and -2.6 kcal mol-1, respectively.

∆Hi ) T∆Si + ∆Gi VOL. 38, NO. 12, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

(5) 9

3341

(4) (5)

(6)

FIGURE 5. Correlation of the thermodynamic parameters in Table 1 for stepwise sorption of HAs on kaolinite at ionic strength 0.05 M (NaCl) and pH 3.5. Sorption of NACs on solid HAs and of dissolved HAs on kaolinite has two common components: humic acids and water. With indications that the need for dehydration of interacting solid HA sites and hydrophilic solutes is a prerequisite for sorption (11, 12), we can compare the slopes (near 300 K) and intercepts (near -4 kcal mol-1) of plots of eq 5 for NAC sorption on solid HAs (11, 12) with the parameters of Figure 5. Similarity of these slopes and intercepts strongly suggests that kaolinite and HA solute dehydration are responsible for the preponderance of entropy increases on sorption of dissolved HAs by solid aqueous kaolinite and the linear correlation in Figure 5. Kaolinite Solubility and HA Desorption. We concluded from many experiments that kaolinite and kaolinite-HA composites dissolve very slowly in 0.1 M NaOH and that HAs desorb very slowly from kaolinite in 0.1 M NaOH at 35.0 °C. Very strong bonds are formed when HAs are sorbed by kaolinite.

Acknowledgments We thank Nadeem Ghali, Nichole Smith, and Marcy Vozzella for many contributions. Staatsbad Pyrmont AG provided the German peat source, the Turf Board Co., Cork, sent the Irish peat sample, Ronald Willey assisted in the kaolinite surface area measurements, Jurgen Liias showed us the source of NHA, and the Johnson family provided the NYHA source from their organic farm in Pine Island, NY. M.E.G., K.O., and T.L.S. were undergraduate research participants.

Supporting Information Available Figures S1 (calibration plots) and S2 and S3 (isotherms) and Tables S1 (analytical data), S2 (13CPMAS data), and S3 (Langmuir parameters). This material is available free of charge via the Internet at http://pubs.acs.org.

(7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19)

(20) (21) (22) (23) (24) (25) (26)

Literature Cited (1) MacCarthy, P. In Humic Substances: Structures, Models and Functions; Ghabbour, E. A.; Davies, G., Eds.; Royal Society of Chemistry: Cambridge, U.K., 2001; pp 19-30. (2) Davies, G.; Fataftah, A.; Cherkasskiy, A.; Ghabbour, E. A.; Radwan, A.; Jansen, S. A.; Kollar, S.; Paciolla, M. D.; Sein, L. T., Jr.; Buermann, W.; Balasubramanian, M.; Budnick, J.; Xing B. J. Chem. Soc., Dalton Trans. 1997, 4047. (3) (a) Sparks, D. L. Environmental Soil Chemistry; Academic Press: San Diego, 1995. (b) Stevenson, F. J. Humus Chemistry:

3342

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 12, 2004

(27)

Genesis, Composition, Reactions, 2nd ed.; Wiley: New York, 1994. (a) Schulten, H.-R.; Schnitzer, M. Naturwissenschaften 1995, 82, 487. (b) Schulten, H.-R.; Schnitzer, M. Soil Sci. 1997, 162, 115 and references therein. (a) Wershaw, R. L. J. Contam. Hydrol. 1986, 1, 29. (b) Wershaw, R. L. Environ. Sci. Technol. 1993, 27, 814. (c) Wershaw, R. L. Membrane-micelle model for humus in soils and sediments and its relation to humification; U.S. Geological Survey Water-Supply Paper 2410; U.S. Geological Survey: Washington, DC, 1994. (d) Wershaw, R. L.; Llaguno, E. C.; Leenheer, J. A. Colloids Surf. 1996, 108, 213 and references therein. Tremblay, L.; Kohl, S. D.; Rice, J. A.; Gagne, J. P. Sorption of PAHs to Natural Sorbents: Impacts of Humic and Lipid Fractions. In Humic Substances: Nature’s Most Versatile Materials; Ghabbour, E. A., Davies, G., Eds.; Taylor & Francis: New York, 2003; p 119. Kretzschmar, R.; Hesterberg, D.; Sticher, H. Soil Sci. Am. J. 1997, 61, 101. Zhou, J. L.; Rowland, S.; Mantoura, R. F.; Braven, J. Water Res. 1994, 28, 571. Maurice, P. A. In Structure and Surface Reactions of Soil Particles; Huang, P. M., Senesi, N., Buffle, J., Eds.; Wiley: New York, 1998; p 109. Chen, Y. In ref 9, p 155. Khairy, A. H.; Davies, G.; Ibrahim, H. Z.; Ghabbour, E. A. J. Phys. Chem. 1997, 101B, 3228. Ghabbour, E. A.; Davies, G.; Fataftah, A.; Ghali, N. K.; Goodwillie, M. E.; Jansen, S. A.; Smith, N. A. J. Phys. Chem. 1997, 101B, 8468 and references therein. Davies, G.; Ghabbour, E. A.; Ghali; N. K.; Mulligan, M. D. Can. J. Soil Sci. 2001, 81, 331. (a) Namjesnik-Dejanovic, K.; Maurice, P. A. Geochim. Cosmochim. Acta 2001, 65, 1047. (b) See also: Namjesnik-Dejanovic, K.; Maurice, P. A. Soil Sci. 2000, 165, 545. Vinodgopal, K.; Subramanian, V.; Carrasquillo, S.; Kamat, P. V. Environ. Sci. Technol. 2003, 37, 761. Pierce, R. H., Jr.; Felbeck, G. T., Jr. Proc. Int. Meet. Humic Subst., Nieuwersluis 1972, 217. Scheffer, F.; Ziechmann, W.; Pawelke, G. Z. Pflanzenernaehr. Bodenk. 1960, 90, 58. Ghabbour, E. A.; Khairy, A. H.; Cheney, D. P.; Gross, V.; Davies, G.; Gilbert, T. R.; Zhang, X. J. Appl. Phycol. 1994, 6, 459. Whyte, J. N. C. Extraction of alginic acid from a brown seaweed. In Experimental phycology: A laboratory manual; Lobban, C. S., Chapman D. J., Kremer, B. P., Eds.; Cambridge University Press: New York, 1988; pp 168-173. Jackson, M. L. Soil Chemical Analysis-Advanced Course; University of Wisconsin: Madison, WI, 1956. ASTM D3663-93. See also: Faeth, P. A. Adsorption and Vacuum Technique: University of Michigan Institute of Science and Technology, 1962; 66100-2-x, 58 pp. Avena, M. J.; Wilkinson, K. J. Environ. Sci. Technol. 2002, 36, 5100. Adamson, A. W. Physical Chemistry of Surfaces, 4th ed.; Wiley: New York, 1982; p 371. Grim, R. E. Clay Minerology, 2nd. ed.; McGraw-Hill: New York, 1968. Mao, J. D.; Hu, W.-G.; Schmidt-Rohr, K.; Davies, G.; Ghabbour, E. A.; Xing, B. Soil Sci. Am. J. 2000, 64, 873. (a) Leenheer, J. A.; Wershaw, R. L.; Brown, G. K.; Reddy, M. M. Appl. Geochem. 2003, 18, 471. (b) Wershaw, R. L.; Llaguno, E. C.; Leenheer, J. A.; Sperline, R. P.; Song, Y. Colloids Surf., A 1996, 108, 199. Rice, J. A. Soil Sci. 2001, 166, 848. See also: Rice, J. A.; MacCarthy, P. Environ. Sci. Technol. 1990, 24, 1875.

Received for review October 29, 2003. Revised manuscript received February 26, 2004. Accepted March 26, 2004. ES0352101