JEQ Journal of Natural Resources and Life Sciences Education
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 Related articles in JEQ
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 ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hur, J.
Right arrow Articles by Schlautman, M. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hur, J.
Right arrow Articles by Schlautman, M. A.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Hur, J.
Right arrow Articles by Schlautman, M. A.
Related Collections
Right arrow Humic Substances
Right arrow Other Contaminants
Right arrow Organic Compounds
Right arrow Soil Pollution
Published in J. Environ. Qual. 33:1733-1742 (2004).
© ASA, CSSA, SSSA
677 S. Segoe Rd., Madison, WI 53711 USA

TECHNICAL REPORTS

Organic Compounds in the Environment

Effects of Mineral Surfaces on Pyrene Partitioning to Well-Characterized Humic Substances

Jin Hura,b and Mark A. Schlautmana,b,*

a Department of Environmental Engineering & Science and Department of Geological Sciences, School of the Environment, Clemson University, Clemson, SC 29634-0919
b Institute of Environmental Toxicology, Clemson University, Pendleton, SC 29670

* Corresponding author (mschlau{at}clemson.edu).

Received for publication July 25, 2003.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Mineral surfaces can alter the ability of humic substances (HS) to bind hydrophobic organic contaminants. In this study, complete adsorption (i.e., to avoid HS adsorptive fractionation effects) of a small subset of well-characterized terrestrial and aquatic HS on kaolinite and hematite significantly changed their subsequent organic carbon–normalized partition coefficients for pyrene relative to their original respective dissolved organic carbon–normalized partition coefficients . Parallel experiments with ultrafiltration (UF) fractions obtained from purified Aldrich humic acid (PAHA) (Aldrich Chemical, Milwaukee, WI) gave similar results. The heterogeneity among the PAHA UF fractions was examined via their mineral surface adsorption characteristics and their subsequent ability to bind pyrene. As expected, variations in maximum adsorption densities (qmax), Langmuir adsorption constants (Kq), and pyrene Kadsoc values were observed among the PAHA UF fractions. However, general trends of qmax, Kq, and pyrene log Kadsoc values for the PAHA UF fractions versus the logarithm of their weight-average molecular weights (MWw) did not typically match the corresponding trends obtained with the four aquatic and terrestrial HS. In general, an ideal mixture competitive adsorption model gave reasonable predictions for PAHA sorption to kaolinite and hematite based on their corresponding UF isotherm parameters. Ideal mixture predictions of pyrene partitioning to adsorbed PAHA from the corresponding UF fraction results were better for kaolinite versus hematite, indicating that the underlying mineral surface can alter the effects of HS heterogeneity on hydrophobic organic contaminant sorption.

Abbreviations: DOC, dissolved organic carbon • HOC, hydrophobic organic contaminant • HS, humic substances • MWw, weight-average molecular weight • PAHA, purified Aldrich humic acid • SHA, soil humic acid • SRFA, Suwannee River fulvic acid • SRHA, Suwannee River humic acid • SUVA, specific ultraviolet absorbance • UF, ultrafiltration


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
THE TRANSPORT, REACTIVITY, bioavailability, environmental effects, and ultimate fate of hydrophobic organic contaminants (HOCs) in biogeochemical systems are greatly affected by the presence of HS (McCarty et al., 1981; Chiou et al., 1983; Karickhoff, 1984). Suspended solid particles and mineral surfaces further complicate these complex systems because a fraction of HS adsorbs at particle mineral surfaces, and the adsorbed and nonadsorbed HS fractions regulate the distribution of HOCs in different ways. Association of HOCs with immobile, adsorbed HS is known to retard their transport, whereas their interactions with mobile, dissolved HS enhance apparent HOC solubilities and thus facilitate their transport in aquatic environments (McCarthy and Zachara, 1989; Johnson and Amy, 1995). The extent of HOC binding to adsorbed and nonadsorbed HS fractions is typically represented by organic carbon–normalized binding coefficients (Koc), which are critical factors in determining HOC distributions. Collective results from previous studies of HOC sorption on natural sorbents reveal that the organic carbon content of sorbents and HOC hydrophobicity are important factors in determining the order of the magnitude of Koc values (Means et al., 1980; Chiou et al., 1983). Variations in Koc values for a given HOC have been attributed to different HS physicochemical properties (Karickhoff, 1984; Gauthier et al., 1987; Chin et al., 1997; Kopinke et al., 2001).

The physicochemical characteristics of different HS vary depending on a multitude of biogeochemical and environmental factors (Thurman, 1985; Stevenson, 1994). For example, terrestrial HS are known to have a higher carbon content, molecular weight (MW), and percentage of aromatic carbon compared with aquatic HS (Stevenson, 1994). Variations in structural and chemical properties comparable with those observed for different HS have also been reported for components within a single bulk HS by elemental analyses and spectroscopic measurements of different fractions derived from one source HS (Shin et al., 1999; Tombácz, 1999; Christl et al., 2000; Hur and Schlautman, 2003a; Khalaf et al., 2003).

Previous HOC sorption studies have used natural soils and sediments collected from various sources (e.g., Means et al., 1980; Garbarini and Lion, 1985; Weber et al., 1992; Huang et al., 1997; Xia and Pignatello, 2001; Gunasekara and Xing, 2003) or minerals coated with HS derived from different sources (e.g., Garbarini and Lion, 1985; Murphy et al., 1990; Schlautman and Morgan, 1993; Onken and Traina, 1997; Laor et al., 1998; Jones and Tiller, 1999; Terashima et al., 2003) to evaluate HOC sorption onto geosorbents. With the latter type of studies, significant differences in HOC sorption have often been reported for the mineral-associated HS versus their corresponding original dissolved HS forms. These differences generally have been attributed to HS conformational changes resulting from their adsorption to mineral surfaces (e.g., Garbarini and Lion, 1985; Murphy et al., 1990; Schlautman and Morgan, 1993; Laor et al., 1998; Jones and Tiller, 1999) and/or HS adsorptive fractionation brought about by the preferential sorption of particular HS components (e.g., Garbarini and Lion, 1985; Laor et al., 1998; Jones and Tiller, 1999). Recent interest in HS adsorptive fractionation with respect to MW (Davis and Gloor, 1981; Wang et al., 1997; Zhou et al., 2001; Johnson et al., 2002; Maurice et al., 2002; Hur and Schlautman, 2003b) suggests the need to better understand variations in the physicochemical characteristics of HS components within a bulk HS material and their subsequent impact on HOC binding reactivities. For example, Hur and Schlautman (2003a) recently demonstrated that selected operational descriptors of HS component properties (e.g., specific ultraviolet absorbance [SUVA], MW) and reactivities (e.g., pyrene log Koc values) could be used to examine the heterogeneity among UF fractions of a purified dissolved terrestrial humic acid (PAHA). Upon comparing the ranges of these properties and reactivities among the PAHA UF fractions to those observed among a small, well-characterized subset of dissolved terrestrial and aquatic HS that included PAHA, they found smaller degrees of variation and different correlations among the various measured parameters. In addition, they applied an ideal mixture model to the measured PAHA UF fraction properties and reactivities and successfully predicted some operational descriptors of those properties (i.e., SUVA, MWw).

The objectives of the present study were to (i) investigate the ability of HS components to bind pyrene while existing in their dissolved versus mineral-associated forms, (ii) examine the sorption behaviors and pyrene Koc values of mineral-bound HS using UF fractions of one source material versus different HS, (iii) investigate the effects of different mineral characteristics on the processes above, and (iv) examine the validity of applying ideal mixture models to HS adsorption on minerals and subsequent pyrene binding. For the above objectives, it is important to note that we wanted to evaluate pyrene binding to the same HS components that differed only in their dissolved versus adsorbed states. Therefore, we were careful to ensure complete HS adsorption to the minerals in this particular phase of the work. Otherwise, such a direct comparison would not have been possible due to the confounding effects resulting from HS adsorptive fractionation.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Materials
The sources, general treatment, and purification of all experimental materials have been previously described (Hur and Schlautman, 2003a, 2003b). Briefly, pyrene was used as a representative HOC. Soil humic acid (SHA), Suwannee River humic acid (SRHA), and Suwannee River fulvic acid (SRFA) were obtained from the International Humic Substances Society (IHSS) and used as received. Purified Aldrich humic acid (PAHA) was prepared by repeated pH adjustment, precipitation, and centrifugation to remove ash, humin, and fulvic acid. Fractionation of PAHA by UF resulted in five different fractions. The relative carbon mass contribution of each PAHA UF fraction (±standard error) was 16.0 ± 0.1, 5.2 ± 0.1, 11.5 ± 0.1, 35.0 ± 0.1, and 32.3 ± 0.1% for the <3 K, 3–10 K, 10–30 K, 30–100 K, and >100 K fractions, respectively (Hur and Schlautman, 2003a). General characteristics of the different HS and PAHA UF fractions are shown in Table 1.


View this table:
[in this window]
[in a new window]
  		Table 1. Specific ultraviolet absorbances (SUVA), organic carbon–normalized pyrene binding coefficients, molecular weights, aromatici-ties, and atomic ratios for the dissolved humic substances (HS) and purified Aldrich humic acid (PAHA) ultrafiltration fractions.{dagger}

 
Kaolinite (Sigma Chemical, St. Louis, MO) and hematite (Alfa; a Johnson Matthey Company, Ward Hill, MA) were used as model soil minerals without additional treatment. Kaolinite is a nonswelling 1:1 layer phyllosilicate clay generally formed from the weathering of granitic rocks and it is a common constituent of many soils (Ko et al., 1998). Hematite ({alpha}-Fe2O3), the second most common iron oxide found in soils, generally occurs in highly weathered soils and gives them their characteristic red color (Sparks, 1995). The overall points of zero charge are often reported to be near 4.2 and 8.2 for kaolinite and hematite, respectively (Ko et al., 1998; Grimaldi, 1999). The specific surface areas for kaolinite and hematite, determined by multipoint N2 adsorption, were 14.3 ± 0.2 and 7.41 ± 0.06 m2/g, respectively (Ko et al., 1998; Grimaldi, 1999).

Analytical Methods
A total organic carbon analyzer (Model 5050; Shimadzu, Kyoto, Japan) was used to quantify the dissolved organic carbon (DOC) concentrations of dissolved HS in aqueous solutions. Dissolved organic carbon concentrations of all samples were kept well above the detection limit (approximately 0.5 mg C/L) in this study, with the exception of HS sorption experiments performed at pH 4 where the objective was to have complete HS adsorption by the minerals to avoid fractionation effects (Hur and Schlautman, 2003b). For the latter case, where residual HS concentrations were below the detection limit of the analyzer, confirmation of the low DOC levels was made by ultraviolet (UV) absorbance. A spectrophotometer (Model DU640; Beckman Instruments, Fullerton, CA) was used to measure UV absorbance of dissolved HS samples. Mineral-preequilibrated aqueous solutions were used as control samples for DOC and UV measurements. Pyrene concentrations in hexane extracts were quantified with a luminescence spectrophotometer (Model LS-5B; PerkinElmer, Wellesley, MA) using external pyrene standards. The excitation/emission wavelengths (nm/nm) for pyrene were 336/373, and the slits were set for bandwidths of 3 nm for excitation and 20 nm for emission. Relative precisions of 1 and 3% were routinely obtained for absorbance/fluorescence and DOC measurements, respectively.

Humic Substance Adsorption Experiments
Adsorption of different HS and PAHA UF fractions onto kaolinite and hematite was examined in batch systems using centrifuge tubes with Teflon-lined screw caps (Pyrex, 50 mL; Corning, Acton, MA). Concentrated mineral suspensions were prepared by adding the minerals to 0.1 M NaCl aqueous solutions and mixing for more than 12 h to allow for full hydration of the surfaces. Aliquots (1 mL) of the concentrated mineral suspensions were added to centrifuge tubes, followed by known concentrations of HS in 0.1 M NaCl aqueous solutions added at different volume ratios. Initial HS concentrations of the samples ranged from zero to a nominal concentration of approximately 23 mg C/L, except for the <3 K PAHA UF fraction for which the highest initial concentration examined was 12.2 mg C/L. Final mineral concentrations were 25 and 5 g/L for kaolinite and hematite, respectively, and the total volume in both systems was 10 mL. All samples were adjusted to a pH of 7, and then placed on a reciprocating shaker for 72 h to equilibrate at room temperature (21 ± 2°C) based on preliminary rate studies. pH measurements after equilibration (pH 6.8–7.1) showed no major differences from their initial value. The mineral particles were separated from solution by centrifugation at 3000 rpm for 30 min. Aliquots (4 mL) of the supernatant were then taken for DOC analysis, and adsorbed HS concentrations were calculated by mass balance. No significant system losses were found in control tests, and all adsorption experiments were conducted in duplicate or triplicate.

Humic substance adsorption isotherms were evaluated using the Langmuir isotherm equation (Tipping, 1981; Gu et al., 1994; Schlautman and Morgan, 1994; Wang et al., 1997; Zhou et al., 2001):

[1]
where q is the adsorption density (mg C/m2 mineral), qmax is the maximum adsorption density (mg C/m2 mineral), C is the equilibrium HS concentration in solution (mg C/L), and Kq is the Langmuir adsorption constant (L/mg C). Nonlinear regression procedures using ORIGIN (Version 6.0; Microcal Software, 1999) were used for fitting Eq. [1] to HS sorption data. It should be noted that good fits of experimental data by the Langmuir equation do not necessarily imply that all assumptions of the model (e.g., homogeneous surfaces, ideal behavior for adsorbed molecules) are fulfilled. For example, Namjesnik-Dejanovic and Maurice (2000) demonstrated using in-solution atomic force microscopy that HS can adsorb in complex structures and aggregates on mineral surfaces.

Determination of Pyrene Organic Carbon–Normalized Partition Coefficient Values for Mineral-Bound Humic Substances
Similarly to the above HS sorption experiments, mineral stock suspensions and aqueous HS stock solutions were prepared separately. Initial HS concentrations were nominally 4 mg C/L. Mineral concentrations were 2.5 and 0.5 g/L for kaolinite and hematite, respectively, and total system volume was the same as for the HS sorption experiments above. The equilibrium pH for these sets of experiments was maintained at 4 so that complete HS adsorption would be achieved, thus simplifying the experimental systems. The presence of nonadsorbed HS, that is, residual HS remaining in solution, would otherwise have led to underestimation of the Koc values for mineral-bound HS by increasing the apparent pyrene solubility (Laor et al., 1998). Moreover, it probably would have resulted in HS adsorptive fractionation (Hur and Schlautman, 2003b). For these particular experiments conducted at pH 4, no appreciable residual HS concentrations were measured in the supernatant solutions after sorption based on UV absorbance measurements at 254 nm. For example, UV absorbance values were always lower than 0.01, which corresponded to a DOC concentration of 0.2 mg C/L, a value below our DOC detection limit. Pyrene stock solutions were then spiked into centrifuge tubes containing appropriate mixtures of the mineral and HS stocks to obtain a total initial pyrene concentration of 100 µg/L. After equilibration (72 h), the solids were separated from aqueous solution by centrifugation at 3000 rpm for 30 min. Aliquots (10 mL) of the supernatant were transferred to 40-mL amber vials. Pyrene in the aqueous phase was quantified by hexane extraction and subsequent fluorescence measurement, generally following the methodology of Johnson and Amy (1995). Briefly, an identical volume of hexane (10 mL) was added to supernatant solutions with the exact volume ratio of aqueous phase and hexane calculated based on gravimetric measurements. The samples were then placed on a high-speed shaker to equilibrate for 1 h. After equilibration, 2 mL of hexane was taken for fluorescence measurement to determine pyrene concentrations based on regression equations developed from external standards in hexane. Preliminary studies showed excellent pyrene recoveries (i.e., 99 ± 2%) in hexane. The quantity of pyrene adsorbed to the solid phase was determined by subtraction of the aqueous pyrene concentration from the known initial concentration. All experiments were conducted in triplicate. Control experiments revealed that 10 ± 1% of the initial pyrene added to the centrifuge tubes adsorbed to the centrifuge tube walls, and this solution removal process was taken into account when calculating the extent of pyrene binding to adsorbed HS. The equilibrium time for pyrene sorption was chosen to be the same as that for HS sorption (72 h) based on preliminary studies.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Sorption of Humic Substances onto Minerals
Sorption isotherms for the different HS and PAHA UF fractions on kaolinite and hematite are depicted in Fig. 1. Although only a few data points are present at the lowest equilibrium concentrations, the initial sharp increase in sorption with equilibrium concentration followed by a leveling off (i.e., saturation) as HS concentrations increase can clearly be seen. All experimental data were reasonably well fit by the Langmuir adsorption model (Fig. 1; Table 2). For some hematite systems, adsorption did not appear to reach saturation at the highest equilibrium concentration investigated, possibly an indication that a change in adsorption affinity due to surface modification by adsorbed HS was operative (Murphy et al., 1990). Differences in the adsorption behaviors of the PAHA UF fractions suggest that important physicochemical characteristics of the fractions differ from one another.



View larger version (32K):
[in this window]
[in a new window]
  		Fig. 1. Humic substance (HS) adsorption isotherms on minerals: (a) different HS on kaolinite, (b) different HS on hematite, (c) purified Aldrich humic acid (PAHA) ultrafiltration (UF) fractions on kaolinite, and (d) PAHA UF fractions on hematite. Adsorption experiments were performed at pH 7 and 0.1 M NaCl in duplicate or triplicate. Error bars for some data points are smaller than the symbols. The solid lines represent the respective Langmuir model fits to the data.

 

View this table:
[in this window]
[in a new window]
  		Table 2. Langmuir isotherm parameters for sorption of purified Aldrich humic acid (PAHA) ultrafiltration (UF) fractions and different humic substances (HS) to kaolinite and hematite.{dagger}

 
Trends of the maximum adsorption densities and Langmuir adsorption constants versus log MWw for the PAHA UF fractions were compared with those for the different HS (Fig. 2 and 3). On kaolinite, the PAHA UF fraction trends for the two adsorption model parameters approximated those of the different HS, although the different HS trends exhibited slightly higher slopes. The observed positive correlations between qmax values and log MWw are consistent with previous studies (Murphy et al., 1990; Balcke et al., 2002), and suggest that MWw–controlling sorption processes such as hydrophobic interactions are predominantly operative for HS sorption to kaolinite (Hur and Schlautman, 2003b).



View larger version (14K):
[in this window]
[in a new window]
  		Fig. 2. Correlations between Langmuir adsorption parameters (pH 7, 0.1 M NaCl) and log weight-average molecular weight (MWw) for purified Aldrich humic acid (PAHA) ultrafiltration (UF) fractions and different humic substances (HS) on kaolinite: (a) maximum adsorption density and (b) Langmuir adsorption constant. Error bars for some data points are smaller than the symbols.

 


View larger version (14K):
[in this window]
[in a new window]
  		Fig. 3. Correlations between Langmuir adsorption parameters (pH 7, 0.1 M NaCl) and log weight-average molecular weight (MWw) for purified Aldrich humic acid (PAHA) ultrafiltration (UF) fractions and different humic substances (HS) on hematite: (a) maximum adsorption density and (b) Langmuir adsorption constant. Error bars for some data points are smaller than the symbols.

 
In contrast to the kaolinite systems, marked differences between qmax and Kq trends for PAHA UF fractions and the different HS were observed with hematite (Fig. 3). For example, a significant positive trend of qmax with log MWw was observed for the different HS but not for the PAHA UF fractions (Fig. 3a). In addition, although a positive trend of Kq values with log MWw was observed for the PAHA UF fractions, a negative trend was found for the different HS (Fig. 3b). This inconsistency between trends for hematite suggests that heterogeneity effects on the sorption processes of the PAHA UF fractions differ from those of the different HS, and also indicates that MWw cannot be a common descriptor for sorption to hematite. The lack of a strong positive trend for qmax with log MWw for the PAHA UF fractions indicates that hydrophobic interactions are not the sole, and perhaps not even the dominant, sorption mechanism operative. For example, the relatively high qmax value observed for the smallest UF fraction suggests that ligand exchange and/or electrostatic attractions are strongly involved because the smaller fractions have more carboxylic groups (Shin et al., 1999). This speculation is supported by the results of Hur and Schlautman (2003b), who quantified the MWw of residual PAHA components in solution after sorption using size exclusion chromatography and demonstrated preferential adsorption of the smaller MWw fractions of PAHA onto hematite under the same solution conditions used here. Additional details on operative HS sorption mechanisms for kaolinite and hematite can be found in Hur and Schlautman (2003b) and references contained therein.

Pyrene Organic Carbon–Normalized Partition Coefficient Determinations for Mineral-Bound Humic Substances
Caution must be exercised in quantifying pyrene Koc values by mineral-bound HS , because some of the adsorbed pyrene may be associated with the bare (i.e., not organically coated) mineral surface (Huang et al., 1996). Mineral contributions to HOC sorption by geosorbents previously have been reported based on observations that simple Koc–Kow partitioning models are not applicable for sorbents with a low organic carbon content (Schwarzenbach and Westall, 1981). McCarty et al. (1981) and Karickhoff (1984) proposed models in which overall HOC sorption consisted of sorption to both the organic and bare mineral phases. Based on their models, the following equation can be used to estimate pyrene Kadsoc:

[2]
where Kd is the observed distribution coefficient (L/g), [PYR]aq is the aqueous-phase pyrene concentration (µg/L), [PYR]solid is the sorbed pyrene concentration (µg/g), Kio is the adsorption coefficient of pyrene on inorganic surfaces (L/g), fio is the fraction of inorganic sorbent surface area available for HOC sorption, and foc is the organic carbon fraction of the sorbent. Note that Eq. [2] is strictly only valid when no solubilizing organic matter is present in the dissolved phase, which is consistent with our experimental system. To use Eq. [2] for determining pyrene Kadsoc values, its Kio values for kaolinite and hematite were measured first in separate sorption isotherm experiments in the absence of HS.

In the presence of HS, the extent of HOC sorption to a bare mineral surface could reasonably be expected to decrease with increasing foc. Although foc values have been equated with fractional surface coverage of adsorbed HS in several literature reports (Murphy et al., 1990; Jones and Tiller, 1999), determining actual fractional surface coverages by adsorbed HS remains problematic. For example, it is likely that HS sorption on kaolinite occurs predominantly at edge sites, whereas for hematite it probably occurs more uniformly across the entire surface (Karickhoff, 1984; Murphy et al., 1990). It is the hope that atomic force microscopy studies may ultimately help to better understand this phenomenon (Maurice and Namjesnik-Dejanovic, 1999; Namjesnik-Dejanovic and Maurice, 2000). In the present study, only minor fractional coverage effects of HS on Kio were assumed (i.e., fio = 1) because the adsorbed HS amounts were relatively low (foc < 0.001). In addition, Murphy et al. (1990) demonstrated that considering HS fractional coverage for HOC sorption to bare mineral surfaces did not markedly affect overall Kadsoc values.

Pyrene binding coefficients by the four different dissolved HS, denoted here by Kdisoc, are compared with their respective Kadsoc values in Fig. 4a. Because residual dissolved HS fractions were largely absent, no adsorptive fractionation occurred in this study. Therefore, Kadsoc values should have approximated Kdisoc if HS conformational changes resulting from their adsorption did not affect pyrene binding. This is an important point, because most previous investigations did not eliminate or rigorously account for HS adsorptive fractionation effects when evaluating their Kadsoc values (e.g., Murphy et al., 1990; Schlautman and Morgan, 1993; Terashima et al., 2003). It can be seen here that mineral-bound HS exhibited higher or lower pyrene binding affinities versus their original dissolved forms (Fig. 4a), and that these differences depended on the specific HS and mineral used. For example, both SHA and PAHA gave higher Kadsoc values for hematite and kaolinite versus Kdisoc, whereas Kadsoc values for SRHA and SRFA were always lower than their respective Kdisoc values. With the PAHA UF fractions, significant differences were also generally observed between Kadsoc and Kdisoc values (Fig. 4b). It can be seen that the terrestrial HS exhibited the trend Kadsoc > Kadsoc > Kdisoc, whereas the trend for aquatic HS was reversed (Fig. 4a). Although the same consistent trend among all three Koc values was not always observed with the terrestrial PAHA UF fractions, the trend Kadsoc > Kadsoc was preserved (Fig. 4b).



View larger version (36K):
[in this window]
[in a new window]
  		Fig. 4. Comparison of pyrene Koc values at pH 4 and 0.1 M NaCl for dissolved and mineral-bound humic substances (HS): (a) different HS and (b) purified Aldrich humic acid (PAHA) ultrafiltration (UF) fractions. Dissolved HS data are from Hur and Schlautman (2003a).

 
Despite the differences between Kadsoc and Kdisoc for the different HS and PAHA UF fractions at pH 4, general trends of increasing log Kadsoc with log MWw similar to those previously reported with the dissolved HS (Hur and Schlautman, 2003a) are readily apparent (Fig. 5), indicating that HS MWw was still controlling the extent of pyrene binding in the foc range investigated despite any conformational changes that occurred on adsorption. In general, the results presented in Fig. 4 and 5 for the different HS at pH 4 are consistent with four previous reports. Murphy et al. (1990) studied anthracene sorption to mineral-bound HS and observed higher Kadsoc values with an IHSS peat humic acid and lower values with SRHA at relatively low foc ranges (<0.001) for both kaolinite and hematite versus values obtained from log Koc–log Kow predictions. Schlautman and Morgan (1993) reported lower perylene Kadsoc versus measured Kdisoc values at various solution chemistry conditions with SRHA and SRFA on an aluminum oxide surface, which would be expected to provide results similar to hematite. Jones and Tiller (1999) used SHA contacted with kaolinite at pH 4 and reported decreasing Kadsoc values with increasing foc. At one very low foc condition, they observed a Kadsoc value significantly higher than that determined for Kdisoc. At foc values comparable with those used in the present study, however, differences between the adsorbed and dissolved partition constants became much smaller, similar to our observations (Fig. 4a and 5a). Recently, Terashima et al. (2003) reported that pyrene binding by terrestrial peat- and coal-derived HS at pH 5 was enhanced by their adsorption to kaolinite. In general, they observed pyrene Koc values that were 4 to 9 times higher with the mineral-bound HS (at the low foc range of 0.0002 to 0.0008) versus the original dissolved HS.



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 5. Correlations between pyrene log Koc values (pH 4, 0.1 M NaCl) and log MWw for dissolved and mineral-bound ({circ}, •: kaolinite; {square}, {blacksquare}: hematite) humic substances (HS): (a) different HS and (b) purified Aldrich humic acid (PAHA) ultrafiltration (UF) fractions.

 
Enhancement in Koc for adsorbed HS thus appears to be limited to relatively large and hydrophobic terrestrial HS, because larger Kadsoc values for both minerals versus their corresponding Kdisoc were observed only with SHA and PAHA, the two larger and more hydrophobic materials investigated. In addition, although the PAHA Kadsoc value for hematite was much higher than Kdisoc, the relative difference became much smaller for kaolinite. Similar to SHA and PAHA, noticeably higher Kadsoc values were observed for hematite versus kaolinite with the larger and more hydrophobic PAHA UF fractions (Fig. 4b). This was despite the fact that a higher foc range was used in the hematite systems. In addition, a slightly greater dependency of Kadsoc on MWw was observed with hematite versus kaolinite (Fig. 5b), which may suggest that pyrene binding sites are hindered to a greater extent for adsorbed PAHA fractions on kaolinite (in particular, for lower MWw fractions). Alternatively, it may be that the different foc levels used for the two mineral systems are responsible for the different dependency observed with kaolinite- versus hematite-bound HS.

It is important to note that the Kadsoc values for PAHA UF fractions do not vary as greatly as for the different HS for both minerals under comparable MWw ranges (i.e., compare reported slopes in Fig. 5). This observation suggests that the extent of pyrene binding is affected not only by physical structural features, but also by HS source and chemical structure even when bound to minerals. In addition, negative correlations between SUVA and Kadsoc values for PAHA UF fractions were observed (data not shown), consistent with the results of dissolved UF fractions (Hur and Schlautman, 2003a). This finding suggests that nonaromatic structural features may be controlling the extent of binding even for adsorbed HS. Collectively, the results above demonstrate that consideration of which HS components are preferentially adsorbed onto minerals (i.e., HS adsorptive fractionation) will probably be critical in understanding HOC sorption by organically coated mineral surfaces.

Ideal Mixture Approach
In a previous report, we examined the feasibility of applying ideal mixture reactivity models to predict overall dissolved PAHA properties and reactivities from its dissolved UF fractions (Hur and Schlautman, 2003a). Here, we extend this concept to test ideal mixture models for PAHA adsorption to kaolinite and hematite and pyrene binding by mineral-associated PAHA in the absence of adsorptive fractionation. If the components within the bulk PAHA material act ideally during sorption and the mineral surface characteristics remain reasonably uniform, then PAHA sorption should be predictable using an ideal competitive sorption model because of the limited number of surface sites available for sorption. Because the sorption isotherms for each PAHA UF fraction were fitted with the Langmuir equation, it was most practical to use a simple competitive Langmuir model to predict the bulk PAHA adsorption. In this case, the qi of the ith component from a bulk HS can be expressed by the following equation:

[3]
where qmax,i and Kq,i are the Langmuir parameters determined from the sorption isotherm of each PAHA UF fraction. Although competition for sorption onto mineral surfaces undoubtedly occurs among the components of a bulk HS, we also evaluated a noncompetitive Langmuir adsorption model for which each PAHA UF fraction was assumed to adsorb independently.

Bulk PAHA adsorption isotherms predicted by the ideal competitive and noncompetitive Langmuir models are shown in Fig. 6 along with the actual experimental data. As expected, the ideal competitive Langmuir model better predicts the actual sorption data for both minerals. The slight differences between the actual data and simulated competitive sorption results probably can be explained merely by experimental uncertainties.



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 6. Comparison of measured and predicted purified Aldrich humic acid (PAHA) sorption isotherms on kaolinite (top) and hematite (bottom). The predicted isotherms were calculated based on ideal competitive and noncompetitive Langmuir adsorption models using the Langmuir adsorption parameters in Table 2 for each PAHA ultrafiltration (UF) fraction.

 
If a heterogeneous mixture of HS components is completely adsorbed to a mineral surface and behaves ideally with respect to pyrene partitioning, the following general mass-balance equation can be used to predict bulk PAHA Kadsoc values from the corresponding values measured for each UF fraction:

[4]
where Kadsoc,i and foci represent the Koc value and relative organic carbon distribution, respectively, of each PAHA UF fraction. Comparison of measured to predicted Kadsoc values shows contradictory results depending on the type of mineral used (Table 3): the predicted Kadsoc is similar to the measured value for kaolinite, but the two values for hematite are statistically different (p = 0.018). However, it should be noted that the percent differences between measured and predicted pyrene Kadsoc values are much smaller than the difference based on dissolved PAHA (Hur and Schlautman, 2003a). Compared with dissolved HS, it may be less surprising that the sum of geometrically or conformationally restricted HS fractions adds up to give the original conformationally restricted total value.


View this table:
[in this window]
[in a new window]
  		Table 3. Comparison of measured and predicted Koc values for pyrene binding by purified Aldrich humic acid (PAHA).{dagger}

 
Collective results from this study and other previous reports (Murphy et al., 1990; Schlautman and Morgan, 1993; Laor et al., 1998; Jones and Tiller, 1999) cast doubt on the accuracy of approaches that attempt to quantify HOC binding by dissolved HS based on subsequent flocculation and precipitation of the HOC–HS complexes using conventional aluminum and iron salts (Laor and Rebhun, 1997). Although the sweep floc resulting from precipitated metal hydroxides is clearly different from the mineral-bound HS systems studied here, it is also obvious that HS conformational changes must occur in those systems as well, which probably perturb the originally established HOC–dissolved HS equilibrium.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
Variations among the UF fractions of PAHA with respect to their sorption onto kaolinite and hematite and their binding of pyrene at the mineral–water interface were observed. In general, trends for these reactivities versus log MWw tended to be different from those established for a small subset of aquatic and terrestrial HS having a comparable MWw range. Positive trends of qmax and Kq with log MWw were observed for PAHA UF fractions on kaolinite. However, the correlation between qmax and log MWw was not significant for hematite, indicating that MWw–controlled sorption processes were not dominating HS sorption to that mineral. Conformational changes of HS on adsorption were probably responsible for the Kadsoc values that were different from their corresponding Kdisoc values, because care was taken to eliminate adsorptive fractionation effects. Nevertheless, adsorbed HS MWw still played a major role in controlling the extent of pyrene binding, as shown by the positive correlations in Fig. 5 between Kadsoc and log MWw for the PAHA UF fractions as well as the different HS. Because the magnitude and direction of differences between Kadsoc and Kdisoc values depended on the particular HS–mineral pairs, an overall general trend showing enhancement of the MWw effect on pyrene binding by mineral-bound HS was observed in Fig. 5. An ideal mixture competitive adsorption model predicted PAHA sorption isotherms reasonably well from its UF fractions for both minerals, and pyrene binding by adsorbed PAHA components appeared to follow more ideal behavior than by their dissolved counterparts. An important caveat regarding our conclusions and data interpretation from this study is that they have been based on macroscopic adsorption and partitioning measurements. Because actual sorption mechanisms cannot be proven by such macroscopic studies alone, it would be extremely beneficial to examine these systems in more detail with complementary spectroscopic and/or microscopic techniques.


   ACKNOWLEDGMENTS
 
We gratefully acknowledge the thoughtful comments and suggestions made by the reviewers and the technical editor, Dr. Dennis L. Corwin. Funding for this work was provided by the National Science Foundation (Grant 9996441) and the USDA (SC-1700133). The contents of this paper do not necessarily reflect the views and policies of NSF or USDA, nor does the mention of trade names or commercial products constitute endorsement or recommendation for use.


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


Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2004 33: 1589-1599. [Full Text]  




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 Related articles in JEQ
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 ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hur, J.
Right arrow Articles by Schlautman, M. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hur, J.
Right arrow Articles by Schlautman, M. A.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Hur, J.
Right arrow Articles by Schlautman, M. A.
Related Collections
Right arrow Humic Substances
Right arrow Other Contaminants
Right arrow Organic Compounds
Right arrow Soil Pollution

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