| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Plant, Soil and Insect Sciences, University of Massachusetts, Amherst, MA 01003
* Corresponding author (bx{at}pssci.umass.edu or bx{at}psis.umass.edu)
Received for publication June 23, 2004.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Abbreviations: CP, cross-polarization • HA, humic acid • HOCs, hydrophobic organic compounds • HS, humic substances • Kd, distribution coefficient at a given concentration • KF, Freundlich coefficient • Koc, organic carbon–normalized sorption coefficient • MAS, magic angle spinning • N, Freundlich exponent or nonlinearity parameter • NMR, nuclear magnetic resonance • OC, organic carbon • SOM, soil organic matter • TOSS, total sideband suppression
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Clay–humic complexes, however, are widely distributed in natural environments. Previous studies have demonstrated that organic matter is likely to be associated with clay fractions as clay–organic complexes in soil and sedimentary environments (McKeague et al., 1986; Ransom et al., 1997). They commonly exist in the form of coatings on solid surfaces (Maurice and Namjesnik-Dejaaovic, 1999). Moreover, Mayer and Xing (2001) have reported that soil samples of low pH, especially from upper soil horizons, have most of their mineral surface covered by organic matter. In this respect, soil organic matter may exert significant impact on the sorption capacity of mineral particles (Zhou et al., 1995) and play a critical role in regulating the retention and transport of contaminants in soils. Even in surface soils containing as little as 1% organic matter, organic components coated on phyllosilicate and oxide minerals were found to dominate the surface chemistry (Bertsch and Seaman, 1999).
Similarly, humic substances (HS or SOM) may undergo chemical or physical changes on adsorption by clay minerals, which may consequently impact their sorption behavior with HOCs (Salloum et al., 2001; Gunasekara and Xing, 2003). Therefore, it may not be appropriate to extrapolate the sorption behavior of HS coated on mineral surfaces from that of extracted HS. Nowadays, a significant research effort has focused on sorption of HOCs either by extracted, nearly 100% organic materials (Chefetz et al., 2000; Xing, 2001b) or by clean minerals (Boyd et al., 2001; Hundal et al., 2001). Information on sorption by humic–clay complexes is rather limited. Murphy et al. (1990)(1994) observed nonlinearity and competitive sorption of three organic compounds by humic-coated minerals as compared with linear and noncompetitive sorption of extracted HAs. Their results indicated that the adsorption mechanism was involved in the uptake process. It has also been reported that partitioning is the dominant process in high organic carbon (OC) content systems while both partitioning and adsorption mechanisms are significant in very low carbon content systems (Onken and Traina, 1997). Moreover, the organic carbon normalized sorption coefficients (Koc) are found to be lower for mineral-bound humic materials compared with dissolved humic materials, inferring HA fractionation during its sorption onto clays (Jones and Tiller, 1999), but no direct evidence has been provided.
With respect to interactions between HS and clay minerals, it has been suggested that high-molecular-weight, hydrophobic fractions are preferentially adsorbed on clay surface (Specht et al., 2000). Nuclear magnetic resonance spectroscopy has recently demonstrated that HA fractions larger than 100000 Daltons are primarily aliphatic in nature, while fractions smaller than 30000 Daltons are predominantly aromatic (Khalaf et al., 2003). Hence, it is reasonable to assume that aliphatic fractions would have higher affinity to clay surface. However, contrary to this assumption, Namjesnik-Dejanovic et al. (2000) observed that more aromatic fractions of a muck fulvic acid were preferentially adsorbed onto clay surfaces. Balcke et al. (2002) also suggested that the sorption affinity to the clay surface was related directly with the aromaticity of HS. Moreover, Gu et al. (1995) reported that carboxyl and hydroxyl groups of HS were actively involved in the adsorption by iron oxide, while from a study on adsorption of HS by kaolinite and montmorillonite, Specht et al. (2000) concluded that the HA fractions with a small apparent molecular size and a high carboxylic group content made up the nonadsorption portion of HS. All these conclusions were inferred from the structural changes of the HS remaining in solution before and after adsorption by clays; the direct characterization of the bound HS by clay minerals has never been reported. Therefore, our objectives for this study were to investigate the structural features of the HA fractions bound to clay minerals and to determine the sorption behavior of those bound HA fractions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Clay–Humic Complex Preparation
A known amount of HA was dissolved in a minimum volume of 0.5 M NaOH to make 500 mg L–1 HA solution after adjusting pH to 5 with 0.1 M NaOH and HCl. Humic–clay complexes were prepared at HA to clay ratios of 1:5, 1:50, and 1:100 (w/w) to obtain different HA loading levels. The HA–clay suspensions were gently shaken on hematology mixers for 24 h and then centrifuged at 7600 x g for 20 min. The precipitates were freeze-dried. The freeze-dried precipitates were washed repeatedly with deionized distilled water until there was no color in the supernatant. The washed precipitate was freeze-dried, ground, and passed a 0.15-mm sieve, and stored for subsequent uses. This procedure was considered to guarantee the short-range reactions between humic substances and clay minerals that would normally have occurred in the soil environment as a result of periodic drying (Varadachari et al., 1991) as well as eliminate the presence of unassociated HA fractions and avoid their interference during structural characterization and sorption experiments. The N2 Brunauer–Emmett–Teller (BET) method with a Beckman Coulter (Fullerton, CA) SA3100 surface area analyzer was used to measure the specific surface areas of the minerals, HA, and clay–humic complexes (Mayer and Xing, 2001).
Carbon-13 Nuclear Magnetic Resonance Spectroscopy
Solid-state 13C NMR techniques were employed to analyze the structural composition of the source HA and the HA fraction bound on clays. The NMR data were acquired at a frequency of 75.48 MHz for carbon on a Bruker (Rheinstetten, Germany) Avance DSX300 spectrometer. The 13C spectrum of Amherst HA was obtained using a total sideband suppression pulse program (cross-polarization magic angle spinning with total sideband suppression, CP/MAS-TOSS) in a 7-mm zirconia rotor with a Kel-F cap. The parameters were: 5-KHz spinning speed, 1.5-ms contact time, 1-s recycle delay, 7000 scans, and 25-Hz line broadening. Only the montmorillonite–humic complex sample prepared at the HA to clay ratio of 1:5 was used for the NMR determination because the carbon contents of other complex samples were too low (Table 1). The 13C spectrum of this complex was acquired using a ramp-cross-polarization pulse program (RAMP-CP/MAS) (Cook and Langford, 1998) with the following parameters: 8-KHz spinning speed, 2.5-ms contact time, 1-s recycle delay, and 114000 scans. The RAMP-CP/MAS technique provides higher signal-to-noise ratio than CP/MAS at a given OC content, and therefore better-resolved spectra can be obtained for low carbon content samples in reasonable experimental time frames. The spectrum of montmorillonite–humic complex sample required 48 h for acquisition. Meanwhile, the empty rotor was also run with the same parameters as the complex sample to acquire a background spectrum.
|
Because there was enough HA remaining in the solution after coating on kaolinite at 1:50 HA to clay ratio, solid-state 13C NMR was used for characterizing the unsorbed fractions, which avoided the concentration limitation and long acquisition time experienced in liquid-state NMR. To ensure the detection of any structural change, the coating procedure was repeated two more times at the same HA to clay ratio. Then, the HA left in the solution was acid-precipitated, centrifuged, washed with deionized distilled water, and freeze-dried. The spectrum of the freeze-dried HA was obtained using CP/MAS-TOSS pulse program as for the source HA. The structural carbon assignments from both solid- and liquid-state NMR were aliphatic carbons (0–50 ppm), methoxyl carbons (50–60 ppm), carbohydrate carbons (60–96 ppm), anomeric carbons (96–108 ppm), aromatic carbons (108–145 ppm), phenolic carbons (145–163 ppm), carboxyl and carbonyl carbons (163–190 ppm), and ketone carbons (190–220 ppm) (Mao et al., 2000; Litvina et al., 2003).
UV-Visible Spectroscopy
The ratio of absorbance of HA at 465 and 665 nm (E4 to E6 ratio) was determined in a sequential coating process to explore the fractionation of HA due to adsorption by clay minerals. This ratio has been used to relate to aromaticity of HA (Stevenson, 1994). In this experiment, montmorillonite or kaolinite was first mixed with 500 mg L–1 HA solution at pH 5.0 with a HA to clay ratio of 1 to 25. Solids were then separated from solution by centrifuging at 7600 x g for 20 min after 24 h of mixing at room temperature, and supernatants were collected. After that, the clean mineral was added to the HA supernatant at approximately the same HA to clay ratio to repeat the coating procedure. Such sequential coating was repeated four times for kaolinite, but only three times for montmorillonite because almost all HA was adsorbed at the fourth coating step. Humic acid solution (or supernatant) was sampled at each coating step and the solution pH was adjusted to 5.0 with diluted HCl and NaOH before measurement. The E4 to E6 ratio of HA was measured with an Agilent (Palo Alto, CA) 8453 UV-visible spectrometer. Because the E4 to E6 ratio is independent of HA concentration (Chen et al., 1977), dilution or other adjustment on HA concentration was unnecessary in this investigation.
Sorption Experiments
The (ring-UL-14C) and unlabeled phenanthrene (>98% purity) were purchased from Sigma-Aldrich (St. Louis, MO) and used without further purification. Sorption experiments were conducted using a batch equilibration method at 20 ± 1°C. The sorption isotherms were obtained using 2-mL glass gas chromatography (GC) vials with Teflon-lined septa cap. The background solution was 0.01 M CaCl2 with 200 µg mL–1 HgCl2 as a biocide. The 14C-labeled phenanthrene and its nonradioactive stock solutions were mixed with humic–clay complexes at different solid to solution ratios. The complex concentrations were adjusted to achieve 30 to 80% sorption of phenanthrene. After the suspensions were shaken for 3 d on hematology mixers to achieve apparent equilibrium, vials were centrifuged at 1000 x g for 30 min, and 1 mL of supernatant was sampled for liquid scintillation counting (Beckman LS 3801). The amount of compound sorbed was calculated by the difference between the applied and equilibrium concentrations. The sorption of phenanthrene by the source HA, kaolinite, and montmorillonite was also performed following the same procedures. Sorption to the GC vials (probably to the Teflon septa caps) was found to account for about 6 to 12% of overall sorption, thus a sorption isotherm was built for GC vials alone and the partition coefficient was used to calculate the amount of phenanthrene sorbed by GC vials in sorption experiments under the assumption of linear partitioning and the similarity of all GC vials. The sorption isotherms for the HA, minerals, and complexes were corrected by subtracting the amount of phenanthrene sorbed by GC vials at each final solution concentration point.
All sorption data were fitted to the logarithmic form of the Freundlich equation:
 |
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Solid and Solution Nuclear Magnetic Resonance Spectroscopy
The solid-state 13C RAMP-CP/MAS spectrum of the montmorillonite–humic complex showed a strong peak at 0 to 50 ppm in the aliphatic region and a large round hump centered between 100 and 200 ppm (Fig. 1A)
. It has been reported that Kel-F rotor caps and NMR probe components may generate significant background signals in aromatic region when samples contain very low OC contents (Preston, 2001). In our case, the carbon content of the complex sample was less than 1%, thus the spectrometer background signals may overlap. Therefore, we ran the empty rotor to acquire background signals at the identical conditions as for the montmorillonite–humic complex. Figure 1B is the spectrum of the empty zirconia rotor with Kel-F cap. Obviously, the round hump in the aromatic region in Fig. 1A was mainly contributed from the spectrometer background signals. The true structural information of the HA fractions adsorbed by montmorillonite (Fig. 1C) was then obtained by free induction decay (FID) subtraction of background signals (Fig. 1B) from the complex sample (Fig. 1A). Hence, the structure of the HA fractions bound to a clay mineral, for the first time, was characterized directly using an NMR technique. It is apparent that the adsorbed HA fractions were primarily aliphatic as compared with the 13C CP/MAS-TOSS spectrum of the source HA (Fig. 1F). In addition, Fig. 1D is the solution 13C NMR spectrum of the HA remaining in the solution after coating on montmorillonite. It provided complementary evidence that aromatic and carboxyl and carbonyl components were dominant fractions of the unsorbed HA. Therefore, we concluded from NMR spectra that aliphatic fractions were preferentially adsorbed on montmorillonite. However, Namjesnik-Dejanovic et al. (2000) reported that aromatic fractions were preferentially adsorbed to mineral surfaces as inferred from absorptivities at the wavelength of 280 nm of a muck fulvic acid in the solution before and after adsorption. This is probably due to the fact that two different humic materials were used in the two studies.
|
|
|
|
|
|
It is anticipated that both HA and minerals could contribute to the sorption of phenanthrene by mineral–HA complexes, and that the higher isotherm linearity might be partially attributed to the mineral phase. To separate the contribution of the bound HA from the overall sorption in the complexes, the sorption contribution from the clay minerals was approximated by assuming that phenanthrene sorption to clays is proportional to surface area and by using the changes in mineral surface area after HA coating (Table 1). Mineral results were subtracted from the total sorption by the complexes to differentiate between mineral and HA coating contribution to sorption. Because it is difficult to determine the conformation and the exact surface area of the sorbed HA on minerals, the surface area of solid HA was used to approximate the HA coverage on minerals. This approximation is reasonable because the surface area of HA of different sources does not seem to vary much (around 1 m2 g–1) (Chiou et al., 1990; Xing and Pignatello, 1997; this study). In addition, though the surface area of HA may depend on its physical size and shape, based on fact that the organic carbon content of clay–HA complexes was so low, a little change on surface area estimation will not affect sorption adjustment.
The adjusted Freundlich parameters of the clay–humic complexes were slightly lower than the unadjusted ones as illustrated in Table 5. Nevertheless, the N values were still higher than the source HA particles. Statistical analysis (two-sample t test) revealed that the N values of montmorillonite–HA complexes prepared at 1:5 and 1:50 and kaolinite–HA complex at 1:5 were significantly different from that of the source HA, while the N values of montmorillonite–HA complex at 1:100 and kaolinite–HA complex at 1:50 were not. According to the dual mode sorption model (Xing, 2001a), these results suggest that the HA bound on the mineral surface might have a less condensed conformation than the source HA particles. The pH at which the complexes were prepared could be one possible reason for the more expanded conformation of the adsorbed HA. In our experiment, the source HA particles were obtained by acidic precipitation (pH 1) while the complexes were prepared at pH 5. Because most functional groups were neutralized at pH 1, HA may have developed a more compacted structure due to intra- and intermolecular attractions. Huang and Weber (1997) also reported that the isotherm N value of a HA was higher at pH 7 than 2.7. Another possible explanation is the fractionation of the HA due to adsorption by mineral surfaces. Previous studies have suggested that the condensed domains and lower N values are often related to aromatic moieties of SOM (Gunasekara et al., 2003; Xing, 2001b; Johnson et al., 2001). Our NMR spectra indicate that a majority of aliphatic components of HA fractions were on mineral surfaces (Fig. 1C), thus generating more linear isotherms than the source HA.
Carbon-normalized distribution coefficients (Koc) were significantly greater with clay–humic complexes as compared with the source HA (Tables 4 and 5). The Koc values of montmorillonite–humic complexes were 60% higher than that of the source HA, while the kaolinite–humic complexes had up to 150% increase in Koc. Again, this increase might be associated with the fractionation of HA on clay mineral surface. According to the 13C NMR (Fig. 1C and 1D), aromatic and carboxyl and carbonyl carbons accounted for the majority of HA left in the solution while the organic carbons on montmorillonite were mainly the paraffinic type of aliphatic carbons (0–50 ppm). Thus, the O to C ratio of the HA fractions on montmorillonite would be lower than that of the HA fractions remaining in the solution and the source HA due to the lack of O-containing functional groups. As a consequence, the adsorbed HA would be more hydrophobic and less polar (Litvina et al., 2003), which favored phenanthrene sorption, resulting in a higher sorption affinity than the source HA. Our recent results and those of other researchers (Kang et al., 2003; Li et al., 2003) demonstrated that aliphatic HA fractions had lower O to C ratios than HAs with more aromatic structures. A similar HA fractionation trend was observed for kaolinite–HA complexes; the HA remaining in the solution after sequential coatings on kaolinite had increased carboxylic and carbonyl carbons and decreased aliphatic carbons (Fig. 1E and Table 2). High sorption affinity with the bound HA fractions tallied with other studies reporting that aliphatic-rich organic samples had high sorption affinity for HOCs (Chefetz et al., 2000; Chefetz, 2003).
The loading level of HA on the mineral surface also impacted sorption linearity and affinity. For both types of clay–humic complexes, N values decreased with decreasing HA loadings suggesting a relatively more condensed character of HA at lower loading levels. Previous study has proposed that the conformation of HA may be rearranged once adsorbed on a mineral surface and the first several molecular layers of amorphous components may take a more condensed form (Gunasekara and Xing, 2003). The Koc values of kaolinite-bound HA appeared to be lower at lower loading levels at both selected equilibrium concentrations (Table 5) while montmorillonite-bound HA exhibited a reversed trend at these two concentrations.
For our experiments, the mineral surface not covered by HA, as estimated by surface area changes before and after coating, was more than 70% for montmorillonite and more than 75% for kaolinite (Table 1). We believe that HA existed in discrete spots on the mineral surface as suggested by Arnarson and Keil (2001). Because different HA loading levels had a similar surface coverage or surface area (Table 1), the HA coating must have increased in thickness while retaining practically the same surface coverage at a higher loading. In such a multilayer arrangement, the first several molecular layers close to the mineral surface may take a more compacted form due to the attractive forces of the mineral surface. This compacted region could enhance nonlinear sorption. Any HA layers beyond this compacted region may be relatively more expanded as the attractive force becomes weaker with distance away from the mineral surface. Therefore, the proportion of the bound HA in the condensed phase (or form) would be relatively high at lower loading, producing isotherms with higher nonlinearity; vice versa for the complexes with higher HA loadings. This may also explain why the montmorillonite–HA complex at 1:100 and kaolinite–HA complex at 1:50 did not have significantly higher N values than the source HA. Therefore, both fractionation and conformation changes contributed to the phenanthrene sorption by the HA fractions associated with minerals.
Although montmorillonite and kaolinite exhibited a similar preferential adsorption for aliphatic HA fractions as demonstrated by NMR spectra, kaolinite–humic complexes had greater Koc values than montmorillonite–humic complexes obtained at the same HA to clay ratio (Table 5). For instance, Koc of the kaolinite–humic complex prepared at 1:5 HA to clay ratio was 80% higher than that for montmorillonite–humic complex at Ce = 0.01 µg mL–1, and 70% higher at 0.1 µg mL–1. Due to the preferential sorption of hydrophobic, aliphatic HA fractions by the clays, the absorbed HA may have a more aliphatic feature in component at lower coating levels at a given HA concentration. Therefore, because montmorillonite adsorbed more than twice as much HA as kaolinite at the same clay to HA ratio, the HA fractions adsorbed on kaolinite would be more aliphatic than that on montmorillonite. Consequently, the HA fractions on kaolinite were relatively more hydrophobic and less polar, resulting in higher phenanthrene sorption on an OC basis.
In summary, NMR data clearly demonstrated that the HA was fractionated during adsorption on mineral surfaces; aliphatic fractions were preferentially adsorbed and aromatic and polar HA fractions were apparently left in the solution. The E4 to E6 ratios of the HA remaining in the solution decreased with further coatings on montmorillonite and kaolinite, indicating a higher aromaticity of HA in the solution, consistent with NMR experiments. For both montmorillonite and kaolinite, complex samples had substantially more linear sorption isotherms and higher sorption affinities than the source HA. The fractionation and conformation changes of HA after adsorption by clays were the main reasons for the elevated sorption (i.e., higher Koc). In addition, the sorption linearity of clay–humic complexes increased with HA loading levels, which may have resulted from a more expanded structure of the bound HA developed at higher loadings. Also, HS and their clay complexes can cause isotherm nonlinearity in soils in addition to black carbons.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Related articles in JEQ:
This article has been cited by other articles:
|
|
 |
|
  Z. He, T. Ohno, F. Wu, D. C. Olk, C. W. Honeycutt, and M. Olanya Capillary Electrophoresis and Fluorescence Excitation-Emission Matrix Spectroscopy for Characterization of Humic Substances Soil Sci. Soc. Am. J., September 1, 2008; 72(5): 1248 - 1255. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Zhang, D. Zhu, and H. Yu Sorption of Aromatic Compounds to Clay Mineral and Model Humic Substance-Clay Complex: Effects of Solute Structure and Exchangeable Cation J. Environ. Qual., May 1, 2008; 37(3): 817 - 823. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Gu and K. G. Karthikeyan Sorption of the antibiotic tetracycline to humic-mineral complexes. J. Environ. Qual., March 1, 2008; 37(2): 704 - 711. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Mikutta and C. Mikutta Stabilization of Organic Matter at Micropores (<2 nm) in Acid Forest Subsoils Soil Sci. Soc. Am. J., October 27, 2006; 70(6): 2049 - 2056. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. He, T. Ohno, B. J. Cade-Menun, M. S. Erich, and C. W. Honeycutt Spectral and Chemical Characterization of Phosphates Associated with Humic Substances Soil Sci. Soc. Am. J., August 22, 2006; 70(5): 1741 - 1751. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. J. Pignatello, Y. Lu, E. J. LeBoeuf, W. Huang, J. Song, and B. Xing Nonlinear and Competitive Sorption of Apolar Compounds in Black Carbon-Free Natural Organic Materials J. Environ. Qual., May 31, 2006; 35(4): 1049 - 1059. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
