SSSAJ Grow Your Career with SSSA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kjaergaard, C.
Right arrow Articles by Villholth, K. G.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Kjaergaard, C.
Right arrow Articles by Villholth, K. G.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Kjaergaard, C.
Right arrow Articles by Villholth, K. G.
Related Collections
Right arrow Batch Studies
Right arrow Colloids
Right arrow Soil Chemistry
Published in Soil Sci. Soc. Am. J. 68:1844-1852 (2004).
© 2004 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA

Division S-2—Soil Chemistry

Properties of Water-Dispersible Colloids from Macropore Deposits and Bulk Horizons of an Agrudalf

Charlotte Kjaergaarda,c,*, Hans Christian B. Hansena, Christian B. Kocha and Karen G. Villholthb

a Chemistry Dep., The Royal Veterinary and Agricultural Univ., Thorvaldsensvej 40, DK-1871 Frederiksberg C, Denmark
b DHI Water & Environment, Agern Allé 11, DK-2970 Hørsholm, Denmark
c Currently at: Dep. of Agroecology, Danish Institute of Agricultural Sciences, PO Box 50, DK-8830 Tjele, Denmark

* Corresponding author (C.Kjaergaard{at}agrsci.dk)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 {zeta}-Potential and Stability...
 CONCLUSIONS
 REFERENCES
 
Mobility of colloids depends in part on the release from aggregates and the stability in suspension. This study determined the soil dispersibility of the different horizons in a Typic Agrudalf. Water-dispersible colloids (WDC) from bulk horizons and macropore deposits were characterized for mineralogy and physical chemical properties. The effect of solution pH and ionic strength on {zeta}-potential and flocculation behavior was evaluated using dynamic light scattering, and the role of colloid associated organic C (OC) and sesquioxides were elucidated. The soil dispersibility did not reflect the contrasting physicochemical conditions, but was directly correlated with total clay content. Generally, WDC were enriched in OC and sesquioxides. Water-dispersible colloids from the Ap-horizon and from macropore deposits differed markedly from the remaining WDC-fractions due to a significantly higher content of OC (14–35 g kg–1), AlCBD (9.3–10.6 g kg–1) and a much smaller N2–adsorption surface area (14–25 m2 g–1). Treatment with Na2S2O8 for removal of OC increased the surface area by 171–225%, indicating surface coatings of OC. The contribution of OC to the colloidal stability was inferred from: (i) a more negative {zeta}-potential and larger suspension stability of WDC with larger content of OC, and (ii) reduced negative {zeta}-potential as well as suspension stability after OC removal. Large variations were observed in the flocculation behavior for WDC with rather similar mineralogical composition. A two-fold increase of the initial particle diameter occurred at an electric conductivity of 91 µS cm–1 for the least stable colloids and at 1023 µS cm–1 for the most stable and OC-rich colloids. The effect of solution pH on flocculation was significant only at pH below 4.5.

Abbreviations: CBD, citrate–bicarbonate–dithionite • EC, electric conductivity • EM, electrophoretic mobility • MM, macropore matrix • MP, macropore deposits • NOM, natural organic matter • OC, organic carbon • OX, treated with sodium peroxodisulphate • PSD, particle-size distribution • WDC, water-dispersible colloids


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 {zeta}-Potential and Stability...
 CONCLUSIONS
 REFERENCES
 
A LARGE NUMBER of inorganic and organic soil contaminants are associated with colloids. The fate of these contaminants and consequently their impact on the environment are thus strongly dependent on the nature and behavior of the colloids. The ability of mobile colloids to facilitate the transport of strongly sorbing contaminants, such as pesticides, P and heavy metals by acting as contaminant carriers has been documented from several laboratory studies (Grolimund et al., 1996; Seta and Karathanasis, 1997; de Jonge et al., 1998, 2000). In addition, results from field studies have demonstrated that soil colloids and associated contaminants may be released to drainage water in high concentrations during rainfall events (e.g., Villholth et al., 2000; Petersen et al., 2003). Thus, mobilization and transport of colloids increases the risk of leaching of strongly sorbing environmental contaminants generally regarded as immobile.

In the vadose zone, the majority of colloids are associated in aggregates and released following aggregate breakdown (Le Bissonnais, 1996). Physical–chemical dispersion is the ultimate state of breakdown that results in release of colloids as a consequence of expanding double layers and dominating repulsive forces as described by the DLVO theory (Derjaguin and Landau, 1948; Verwey and Overbeek, 1948). It is generally recognized that the fraction of clay that disperses in water (WDC) has been found related to soil erodibility (e.g., Brubaker et al., 1992), and WDC have also been used as an input parameter for predicting colloid leaching through the vadose zone (Jarvis et al., 1999). From regression analysis comparing a range of soil factors, several studies have identified total clay content as one of the most important properties in determining the amount of WDC (e.g., Pojasok and Kay, 1990; Brubaker et al., 1992). Seta and Karathanasis (1996), however, did not find any correlation between total clay content and the amount of WDC investigating subsurface alfisols, mollisols, and ultisols. They argued that the Fe, Al, and kaolinite content accounted for most of the variability in WDC. It is generally recognized that soils dominated by 2:1-layer minerals are dispersed more readily than those dominated by sesquioxides and 1:1-layer minerals (Yaron and Thomas, 1968). The higher dispersibility of soils dominated by 2:1-layer minerals has been attributed to a higher charge density, but Seta and Karathanasis (1996) concluded that the mineralogical composition and surface charge did not explain differences in colloid dispersibility among illitic, montmorillonitic, and mixed Alfisols. In natural field soils, the effect of mineralogy may, however, be influenced by the presence of surface adsorbed OC, which may mask the direct importance of clay mineralogy. Experimental evidence has suggested that natural organic matter (NOM) adsorbed to soil colloids stabilizes colloid suspensions due to both electrostatic and steric mechanisms (e.g., Kretzschmar et al., 1993; Kaplan et al., 1997; Kretzschmar et al., 1998), and organic coatings have been found to induce colloid stability by steric hindrance at high ionic strength conditions (Hunter, 1987).

Once colloids have been dispersed, they can be transported through natural porous media at a greater velocity than conservative dissolved tracers (Kretzschmar et al., 1995), the transportability being determined both by the pore size of the actively conducting flow pathways and the size and stability of the dispersed colloids in the soil solution. When soil water containing suspended colloids infiltrates deeper soil layers with changing chemical environment, continued mobility depends on the ability of colloids to resist flocculation and sedimentation. Identification and quantification of the solution and colloidal properties controlling the stability of the potentially mobile colloids is consequently of importance when predicting the susceptibility for colloid leaching to drainage and ground water.

In Danish clayey till soils where the clay fraction is dominated by 2:1-layer silicates, dispersion and translocation of clay colloids is a natural process and a key phenomenon, resulting in the development of illuvial subsurface horizons with a higher clay content compared with the upper eluvial horizons. Furthermore, micromorphological features showing deposits of clay skins on ped faces and at the interface of water conducting pores represents evidence of colloid translocation. Conducting a micromorphological analysis of a Typic Agrudalf (Flakkebjerg, Denmark), Rasmussen et al. (2001) observed well-developed clay coatings at the interface of fractures, interaggregatepores and biopores. These observations indicated a pronounced colloidal transport to the depth of 170 cm. The objectives of this study were to: (i) determine the soil dispersibility of the different horizons in the soil profile, (ii) determine the mineralogy and physicochemical properties of WDC isolated from macropore deposits and bulk horizons, and (iii) determine the effect of solution ionic strength and pH and evaluate the role of OC on colloid stability based on measurements of {zeta}-potential and particle-size distributions.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 {zeta}-Potential and Stability...
 CONCLUSIONS
 REFERENCES
 
Field Site and Soil Characteristics
The investigated soil originates from a site located on a flat glacial plain at Flakkebjerg, Denmark. The top-till is developed on morainic deposits from the Weichsel Glacial Age, and the site has been cultivated for centuries. The soil profile was excavated to a depth of approximately 150 cm, and bulk soil material was sampled over the entire horizon from each horizon in the profile. Intact structures for preparation of thin sections and colloid characterization were collected around pores with visible clay and organic deposits in the BCtg horizon. Special samples comprising macropore deposits (MP-BCtg) and underlying macropore matrix material (MM-BCtg) were separated using a scalpel. Thin sections were prepared from polyester resin impregnated intact structures (Dalsgaard et al., 1981; Murphy 1986). Thin sections were examined with a petrologic microscope (Carl Zeiss Jena–Jenapol). All bulk soil samples were air-dried, gently crushed, sieved at 2 mm, and used for all analysis. Particle-size distribution of the <2-mm fraction was measured using the Andreasen pipette method (Gee and Or, 2002) after dispersion in sodium pyrophosphate. The content of calcite if present was measured gas volumetrically. Soil pH was measured in water in a 1:8 soil/water suspension; this soil/solution ratio was used as it represents the same soil/solution ratio used for fractionation of WDC. Cation-exchange capacity at pH 8.1 was determined using the ammonium acetate method (Chapman, 1965). Sodium adsorption ratio (SAR) was calculated from the concentrations of Ca2+, Mg2+, and Na+ in the leachate. Iron (FeCBD, Feox) and Al (AlCBD, Alox) were analyzed after extraction by citrate–bicarbonate–dithionite (CBD) (Mehra and Jackson, 1960), and oxalate (ox) (Schwertmann, 1964). The CBD-treatment was repeated twice to ensure complete removal of oxides. All analyses were made in duplicate except for soil texture. Average values are presented in Table 1.


View this table:
[in this window]
[in a new window]
  		Table 1. Chemical and physical properties of the Flakkebjerg profile.{dagger}

 
Fractionation of Water-Dispersible Colloids
The fraction of WDC was separated from the air-dried and sieved bulk samples (Ap, Btg2, and BCtg) and special samples (MP-BCtg and MM-BCtg) by shaking in deionized water at a 1:8 (w/w) soil/water ratio for 16 h on a reciprocal shaker at 60 rpm. After dispersion, the <20-µm fraction was separated and decanted from the soil suspension by gravity sedimentation. Further fractionation of the decanted suspension in particle-size fractions 0.2 to 2 and <0.2 µm was accomplished by centrifugal particle-size analysis (Slater and Cohen, 1962).

Each WDC fraction was divided into three portions: One remained intact, the second was treated once with sodium peroxodisulphate (OX) for oxidative removal of organic matter (Meier and Menegatti, 1997), and the third portion received the Na2S2O8 treatment followed by a treatment with CBD for removal of Fe and Al oxides (Mehra and Jackson, 1960). The treatments are subsequently referred to as intact, OX, and OX + CBD. Oxidation using Na2S2O8 was chosen because it is more effective in removing organic matter from clays than procedures employing H2O2, and because it imparts minimum destruction of the inorganic mineral phase (Meier and Menegatti, 1997). All WDC fractions were subsequently saturated with Ca2+ by five centrifugal washes with 1 M CaCl2. Excess electrolytes were removed by centrifugal washes with deionized water until the electric conductivity (EC) of the supernatant was ≤10 µS cm–1. The WDC-suspensions were stored in polyethylene containers at 4°C until use. The concentration of solids in each colloidal suspension was determined in triplicate as suspension dry weight, after oven drying at 100°C. Subsamples of each WDC fraction were saturated with Mg2+ and K+, using the same procedure as with CaCl2. These samples were subsequently air-dried and used for mineralogical and chemical analysis.

Characterization of Water-Dispersible Colloid Fractions
Iron (FeCBD) and Al (AlCBD) were extracted in triplicate from the intact WDC-fractions by CBD treatment (Mehra and Jackson, 1960), and determined using atomic absorption spectroscopy (AAS). Total C and N were determined in duplicate for intact and OX-treated WDC-fractions using a Europe Scientific mass spectrometer (Europe Scientific, Crewe, UK). The mineralogy of the OX + CBD treated fractions were investigated by powder X-ray diffraction (XRD) using a Siemens D 5000 diffractometer (siemens AG, Karlsruhe, Germany) and applying Co-K{alpha} radiation at a scanning speed of 0.24°2{theta} min–1. Both unoriented and oriented Mg2+, Mg2+ + ethylene-glycol and K+ saturated specimens were examined, the latter additionally at 350 and 550°C. Powder X-ray diffractograms of the colloid fractions were obtained and semi-quantitative estimation of the clay mineral composition was accomplished according to the procedure described by Ernstsen (1998). Specific surface areas were determined through N2–adsorption using a single-point BET method.

Electrophoretic Mobility and Flocculation Behavior
Electrophoretic mobility (EM) and particle-size distributions (PSD) of WDC-suspensions were measured by laser Doppler velocimetry-photon correlation spectroscopy (LDV-PCS) using a Zetasizer3000 instrument (Malvern Instruments, Malvern, England). Measurements of EM and PSD were calibrated using latex standards supplied by Malvern Instruments. The mean hydrodynamic diameter (dH) of colloids and flocs, which by definition is inversely proportional to their diffusion coefficient, is obtained from a cumulant analysis of the autocorrelation function of the scattered light intensity. Zeta potentials ({zeta}) were estimated from EM measurements (µ) using Helmholtz-Smoluchowski's equation (Hiemenz, 1986). All analyzes were performed at 20 ± 0.5°C, and triplicate measurements were recorded for each sample.

Analysis was performed on the intact, OX, and OX + CBD treated WDC-fractions as a function of pH and EC of CaCl2 suspensions. Colloidal suspensions for EM and PSD analysis were prepared from stock suspensions. The stock suspension was diluted in six different CaCl2 concentrations bracketing an EC from 10 to 1200 µS cm–1 to provide colloid mass concentrations of 25 mg L–1 for EM measurements and of 100 mg L–1 for PSD measurements. Colloid stability as a function of pH was measured in a background electrolyte concentration of 0.35 mM CaCl2 (corresponding to an EC of 78 µS cm–1), where pH was adjusted to initial values of 3.5, 4.5, 5.5, 6.5, 7.5, and 8.5 by addition of appropriate amounts of HCl or Ca(OH)2. Suspensions were allowed to equilibrate by shaking on a reciprocating shaker at 60 rpm for 20 h before measurements of either EM or PSD. The final suspension pH and EC was measured immediately before analysis.

Statistical Analysis
Data were analyzed using General Linear Model's procedures for comparison (SAS Institute, 1990). Least significant differences at the 5% level of significance (LSD0.95) were calculated.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
{zeta}-Potential and Stability...
 CONCLUSIONS
 REFERENCES
 
Characterization of the Soil Profile
The Flakkebjerg profile was classified as a loamy, mixed, mesic Typic Agrudalf (Soil Survey Staff, 1992). The profile was well-structured with several vertically oriented macropores penetrating to a depth of 150 cm. Eluviation of clay particles from the top horizon was evident from clay skins on ped surfaces and at pore interfaces, and the enrichment of clay in the Btg and BCtg horizons (Table 1). A micromorphology image of an intact structure showing a typical macropore area with deposits of laminated clay colloids is presented in Fig. 1 .



View larger version (142K):
[in this window]
[in a new window]
  		Fig. 1. Micromorphology image (plane-polarized light) of an intact structure from the BCtg-horizon showing (a) macropore deposits (MP-BCtg) of laminated clay colloids, (b) macropore matrix (MM-BCtg), and (c) macropore area.

 
The parent material was calcareous as indicated by the pH (6.2–8.3) and the presence of calcite (73.3 g kg–1) in the BCtg horizon (Table 1). The CEC8.1 was well correlated with clay content (R2 = 0.997). Calcium was the dominating base cation resulting in very low sodium adsorption ratios. The OC content was highest in the Ap horizon (11.1 g kg–1), but small amounts were found throughout the profile (1.1–4.4 g kg–1). The C/N ratio associated with the organic matter was close to 10 in the Ap horizon, and decreased to approximately 6 in the lower horizons. The largest amount of FeCBD (12.0 g kg–1) and AlCBD (1.17 g kg–1) were found in the Btg1 horizon. The FeOX/FeCBD ratio was in the range 0.10 to 0.47, with the highest value in the Ap horizon exposed to the most intense weathering, while the Fe oxides in the lower soil horizons were more crystalline. Mineralogical analysis of the bulk soil, as investigated by Rasmussen et al. (2001), showed the presence of smectite, illite, interstratified illite-smectite (I-S), vermiculite, hydroxy interlayered vermiculite (HIV), and kaolinite, with smectite as the dominant clay mineral (40–65% of the clay minerals).

Colloid Dispersibility
The percentage of WDC recovered from bulk samples increased down the profile, from 6.8% in the Ap-horizon, 12.7% in the Btg2 horizon, to 17.9% in the BCtg horizon (Table 2). These results demonstrated that from 46% (Btg2 horizon) to more than 50% (Ap and BCtg horizon) of the total soil clay was found in the WDC fraction. The macropore deposits were found to be highly dispersible with 46.5% of the total mass fraction found in the WDC-fraction, compared with only 13.9% in the underlying matrix material (MM-BCtg). The largest amounts of WDC were generally found in the 0.2- to 2-µm size fractions, except for the macropore deposits, which were dominated by fine clay <0.2 µm. We would expect that differences in WDC between horizons would reflect contrasting mineralogical or soil physicochemical properties, but the high amount of WDC in the BCtg horizon despite the presence of calcite, indicated that measurements of WDC did not reflect these characteristics. Instead we found that the mass of WDC in the bulk samples was strongly correlated with the content of total clay for all horizons sampled (R2 = 0.999). This is in agreement with most other studies demonstrating a positive correlation between total clay content and the amount of WDC (e.g., Pojasok and Kay, 1990; Brubaker et al., 1992; Rasiah et al., 1992; Levy et al., 1993; Curtin et al., 1994). Common for these studies is the application of air-dry soil and the use of mechanical shaking in the measurement procedure. Results from Kjaergaard et al. (2004a) also demonstrated a positive correlation between clay content and WDC using air-dry soil and mechanical shaking, while employing a low-energy input separation procedure yielded a negative correlation between WDC and total clay content. Additionally, the low-energy input measurement of WDC was additionally very sensitive to initial soil conditions, and reflected the actual leaching of colloids from intact soil cores (Kjaergaard et al., 2004b). The authors therefore recommended that separations of WDC, that should resemble potentially mobile colloids, should be based on low-energy input separation procedures. Based on this knowledge we expect that part of the separates of WDC obtained in this study and in other studies using intensive mechanical shaking, may represent some colloidal material that are not easily dispersed.


View this table:
[in this window]
[in a new window]
  		Table 2. Physicochemical properties of water-dispersible colloids (WDC) from the Flakkebjerg profile.{dagger}

 
Characterization of Water-Dispersible Colloids
Semiquantitative estimates of the relative content of clay minerals revealed that the mineralogy of the WDC in general were dominated by smectite or smectite + vermiculite, with smaller amounts of interstratified smectite-illite, illite, hydroxy-interlayered smectite/vermiculite, and kaolinite (Table 2). This mineralogy corresponded to that of the total clay fraction of this soil (Rasmussen et al., 2001), indicating that there was no preferential dispersion of any clay mineral phase in the WDC-fractions. This could be attributed to the high-energy input in the WDC-separation procedure yielding WDC-separates constituting about 50% of the total clay fraction. The WDC-fractions showed enrichment in OC, compared with the bulk soil, with WDC from the Ap-horizon having the largest content of OC (26.1–35.3 g kg–1), followed by the MP-BCtg fractions (13.8–17.9 g kg–1). The NOM associated with the WDC from the BCtg2 and BCtg fractions, generally had a low C/N ratio (5.8–6.6), not very different from the C/N ratios of the bulk soil material (6.1), reflecting the same type of OC. In contrast a higher C/N ratio (8.0–8.6) was observed for the WDC from the Ap-horizons, which differed markedly from the C/N ratio of the bulk soil material (9.8), indicating other components of NOM in the WDC-fractions compared with the bulk soil.

Another characteristic feature concerning the composition of the WDC fractions was the enrichment of sesquioxides. The relatively large content of sesquioxides, in particular Al oxides, in the WDC fraction was remarkable. Iron and Al are known to be strong flocculants (Goldberg et al., 1990; Seta and Karathanasis, 1996), and based on the general recognition of the flocculation power of sesquioxides, it cannot be expected that the oxides in the WDC appeared as free or surface exposed components, but rather screened by other components such as surface adsorbed organic matter. In the present case, the results indicate a positive correlation between the WDC content of Al and OC (R2 = 0.770). Sesquioxides are indeed effective sorbents of humic substances, where specific adsorption via ligand exchange with protonated surface hydroxyl groups is the dominating binding mechanism (Murphy and Zachara, 1995). Shen (1999) reported results from scanning electron microscope (SEM) investigations, indicating the existence of a correlation between the atomic ratio of (Al + Fe)/Si of soil colloid surfaces and the humic acid sorption capacity, where those soil colloids with larger values of the (Al+Fe)/Si ratio were able to sorb humic acid more efficiently.

Measurements of N2–BET specific surface area of the WDC fractions (Table 2) showed a variation in specific surface area from 14 to 69 m2 g–1. The lowest specific surface areas were observed for the Ap and MP-BCtg fractions, also having the largest contents of OC. The relatively low N2–BET surface area combined with the large content of OC, indicate that organic matter dominates the surfaces of these WDC-fractions. Treatment with Na2S2O8 removed 68 to 92% of the total OC from these WDC fractions, and resulted in an increase in specific surface area by 171 to 225% (Table 3). This supported the hypothesis that the surfaces of these WDC were at least partly coated by OC, known to have extremely low (<1 m2 g–1) N2–gas adsorbing surface areas (Chiou et al., 1990). Additional treatment with CBD for removal of Fe and Al oxides decreased the specific surface area of the Ap fractions, but had less effect on the surface area of MP-BCtg fractions.


View this table:
[in this window]
[in a new window]
  		Table 3. Effect of removal of organic carbon (OX) and organic carbon + sesquioxides (OX + CBD) on specific surface area and particle diameter of water-dispersible colloids (WDC).

 

   {zeta}-Potential and Stability Behavior of Natural Water-Dispersible Colloids
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 {zeta}-Potential and Stability...
CONCLUSIONS
 REFERENCES
 
Increasing CaCl2 concentration resulted in a marked decrease in the negative {zeta}-potential for all WDC fractions, from about –35 to –45 mV at EC approximately equal to 10 µS cm–1 to about –11 to –15 mV at EC approximately equal to 1200 µS cm–1 (Fig. 2a) . The {zeta}-potential differed between the WDC fractions, with the most negative {zeta}-potentials observed for the Ap fractions, followed by the MP-BCtg fractions. The reduction in negative {zeta}-potential with increasing EC can be explained by the compression of the diffuse double layer around charged surfaces, and a subsequent reduction in the electrical potential at the plane of shear (Hunter, 1981). In addition, increasing Ca2+ concentration in solution may result in increased Ca2+ binding to functional groups of colloid-bound humic substances, and a subsequent charge reduction (Amirbahman and Olson, 1995). The reduction in {zeta}-potential indicates reduction in the repulsive forces, and this was demonstrated in the stability behavior of the WDC. Increasing the CaCl2 concentration resulted in flocculation, as confirmed by the relative increase in the mean hydrodynamic particle diameter (Fig. 2b). The increase in dH differed between the WDC fractions, but the flocculation pattern revealed identical behavior. At low CaCl2 concentrations, the suspensions were stable and no flocculation occurred, but at a certain critical concentration, flocculation was initiated. The onset of flocculation differed among the WDC, and a two-fold increase of the initial particle diameter occurred at an EC of 91 µS cm–1 for the least stable Btg2 colloids, 171 to 185 µS cm–1 for MP-BCtg, and at 1023 µS cm–1 for the most stable Ap colloids. The Btg2 (0.2–2 µm) fraction was very unstable and allowed only measurements at the lowest CaCl2 concentrations before flocculation reached a level where further increase in particle size could not be measured due to sedimentation. The stability behavior of the WDC can, at least partly, be explained by their respective {zeta}-potentials, the colloids having the most negative {zeta}-potential also revealed the largest resistance to flocculation (Fig. 2a,b). However, at higher CaCl2 concentrations when the {zeta}-potential was reduced to a minimum level for all WDC fractions, steric stabilization forces from surface adsorbed organic matter may induce colloid stability at high ionic strength conditions (Hunter, 1987).



View larger version (31K):
[in this window]
[in a new window]
  		Fig. 2. Influence of CaCl2–concentration (EC) and pH on (a, c) {zeta}-potential, and (b, d) relative increase in mean hydrodynamic particle diameter (dH) of water-dispersible colloids (WDC) from different soil horizons and colloidal size fractions. Error bars: ± SE correspond to measurements on three replicate samples.

 
In contrast to the marked influence of solution CaCl2 concentration on the {zeta}-potential and stability behavior, the effect of solution pH was less pronounced. It appeared that the {zeta}-potential for the WDC was almost constant at pH 3.5 to 8; only the Ap (<0.2 µm) showed some reduction in the negative {zeta}-potential at pH < 5 (Fig. 2c). The relative increase in the mean hydrodynamic particle diameter as a function of pH (Fig. 2d) showed that at pH 5 to 7, the colloidal suspensions were stable with no flocculation taking place. At pH < 5 there was a relative increase in dH for all colloid fractions except the Ap (<0.2 µm) reflecting flocculation, and at pH < 4 the destabilization of the colloid suspension was marked for all fractions. Again, the Ap colloids appeared to be more resistant toward flocculation than the MP-BCtg colloids, and the Btg2 colloids were significantly less stable. At pH > 7 there was a slight decrease in dH for some of the WDC indicating dispersion at higher pH. The effect of pH on the stability behavior can be explained as a result of protonization of pH-dependent charges sites on sesquioxides, humic substances, and on the edge of layer silicates, and attraction between these positively charged surface sites and negatively charged layer silicates. Release of Al3+ at low pH may also contribute to flocculation. Furthermore, the changes in EC caused by the pH adjustment when adding HCl and Ca(OH)2 increases the ionic strength of the suspension. Theoretically, we would expect this to additionally decrease the negative {zeta}-potential and increase flocculation at both low and high pH where the rise in ionic strength is largest.

Influence of Natural Organic Matter and Sesquioxides on {zeta}-Potential and Stability Behavior
The {zeta}-potential and stability behavior of the intact WDC fractions indicated that the stabilization of the suspended colloid fractions resulted from surface coverage by organic matter with high charge density. The OX-treatment for removal of NOM from Ap and MP-BCtg fractions had two significant effects on the resulting {zeta}-potential (Fig. 3) . First, removal of NOM markedly reduced the negative {zeta}-potential, and second, removal of NOM caused the {zeta}-potential to become markedly affected by pH. The reduction in {zeta}-potential was more pronounced for the WDC fractions from MP-BCtg (43–54%) than from the Ap horizon (17–26%), but the resulting pH-dependence of the {zeta}-potential showed similar patterns. As pH increased the reduction in {zeta}-potential, due to removal of NOM, became less marked, and at pH 7 through 8 the reduction in {zeta}-potential was only 17 through 18% for the MP-BCtg and 4% for the Ap (<0.2 µm). For the Ap (0.2–2 µm) removal of NOM at pH > 6 resulted in an increase in the negative potential, reflecting an increase in the negative surface charge.



View larger version (25K):
[in this window]
[in a new window]
  		Fig. 3. Effect of removal of organic matter (OX) and organic matter plus sesquioxides (OX + CBD) on the {zeta}-potential of water-dispersible colloids (WDC) from different soil horizons and colloidal size fractions as a function of pH. Error bars: ± SE correspond to measurements on three replicate samples.

 
An explanation for this change in {zeta}-potential versus pH characteristics is that removal of NOM exposes a surface of sesquioxides with pH-dependent charge behavior. At low pH, the oxides will be dominantly positively charged, and therefore contribute to a marked decrease in negative {zeta}-potential, when removing surface adsorbed NOM. As pH increases, the positive charge of the oxides is reduced; hence the contribution of the sesquioxides to change in {zeta}-potential decreases. The anomalous behavior of Ap (0.2–2) at pH > 6 may indicate the presence of sesquioxides with lower point of zero charge (PZC) than for the other samples.

The above considerations are supported by the increase in the negative {zeta}-potential for the OX + CBD treated fractions compared with the OX treated fractions (Fig. 3). Removal of sesquioxides from the Ap fractions increased the {zeta}-potential either to the level of the intact WDC, or to even more negative potentials. For the MP-BCtg the removal of sesquioxides resulted in an increase in {zeta}-potential to an intermediary level for the MP-BCtg (<0.2 µm), while the resulting {zeta}-potential for the MP-BCtg (0.2–2 µm) was increased to the level of the intact fraction at pH ≥ 5. Oxidative removal of NOM from Ap fractions reduced the {zeta}-potential less than removal of NOM from the MP-BCtg fractions, despite the larger content of OC and the more negative {zeta}-potential of the Ap fractions. The reason for this cannot be deduced from this study, but it is likely to be related to the type and morphology of the underlying layer silicate and sesquioxide fraction.

The destabilizing effect of NOM removal on the WDC fractions was reflected in the flocculation behavior (Fig. 4) . The OX-treated colloids rapidly flocculated even at high pH, whereas the suspensions of intact colloids remained stable at pH above 4 and 5. The pH dependence was extremely pronounced for the OX-treated colloids. An exception to this behavior was the Ap (0.2–2 µm) fraction where the NOM removal caused only a minor increase in the particle diameter. This is however in accordance with the less significant effect of NOM removal on the {zeta}-potential for this fraction. Furthermore, removal of sesquioxides from the OX treated fractions resulted in a reduction in the average particle diameter. Smaller particle size of OX + CBD fractions results from loss of sesquioxides as "cement" gluing particles together, and the reduction in particle flocculation when surface coatings of positively charged sesquioxides are removed.



View larger version (26K):
[in this window]
[in a new window]
  		Fig. 4. Effect of removal of organic matter (OX) and organic matter plus sesquioxides (OX + CBD) on the mean hydrodynamic particle diameter (dH) of water-dispersible colloids (WDC) from different soil horizons and colloidal size fractions as a function of pH. Error bars: ± SE correspond to measurements on three replicate samples.

 
The larger stability of the NOM coated colloids suggests that NOM plays a significant role in stabilization of suspended colloids in this soil type dominated by 2:1-layer silicates. This is in agreement with results reported earlier for suspensions of natural variable charged soil colloids (Kretzschmar et al., 1993), and for artificial mixtures between variable charged minerals and humic acids (Jekel, 1986; Kretzschmar et al., 1998). This study demonstrated that for WDC with rather similar mineralogical composition, large variations were observed in the flocculation behavior. These findings clearly demonstrate that model predictions of colloid deposition based on colloid mineralogy may lead to erroneous conclusions. Also for colloid fractions where the stability behavior is controlled by surface adsorbed organic matter, large variation may occur in colloidal stability. This suggests that model predictions of colloid deposition should be based on direct examinations of colloid stability behavior.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 {zeta}-Potential and Stability...
 CONCLUSIONS
REFERENCES
 
This study demonstrated that colloid associated organic matter increased the suspension stability of natural WDCs in a Typic Agrudalf dominated by 2:1-layer silicates. The contribution of OC to the colloidal stability was inferred from: (i) a more negative {zeta}-potential and larger suspension stability from WDC with larger content of OC, and (ii) reduced negative {zeta}-potential as well as suspension stability after OC removal. The presence of surface adsorbed organic matter presumably masked the underlying silicate and sesquioxides components, and consequently controlled the stability behavior of the WDC.

The average {zeta}-potential and stability behavior of natural WDC were strongly dependent on solution CaCl2 concentration. Flocculation of WDC was directly related to {zeta}-potential measurements in the lower concentration range, while at higher CaCl2 concentrations steric stabilization forces may additionally explain the colloidal stability. Solution pH had less effect on the stability of the WDC, and flocculation was only significant at pH below 4.5.

This study demonstrated that a conventionally used method for separating WDCs using air-dry soil and mechanical shaking yielded WDC-fractions that did not reflect differences in soil physical-chemical conditions, but was directly related to total clay content. We suggest that future investigations of WDC should consider separation procedures that may better resemble the easily dispersible colloid fraction.


   ACKNOWLEDGMENTS
 
The authors thank Lars Holm Rasmussen, Chemistry Department, The Royal Veterinary and Agricultural University, for carrying out the thin section analysis.

Received for publication May 20, 2003.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 {zeta}-Potential and Stability...
 CONCLUSIONS
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kjaergaard, C.
Right arrow Articles by Villholth, K. G.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Kjaergaard, C.
Right arrow Articles by Villholth, K. G.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Kjaergaard, C.
Right arrow Articles by Villholth, K. G.
Related Collections
Right arrow Batch Studies
Right arrow Colloids
Right arrow Soil Chemistry

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		 Journal of
Environmental Quality