| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Soil & Crop Sciences Dep., Texas Agricultural Experiment Station, Texas A&M Univ., College Station, TX 77843
b Dep. of Crop and Soil Sciences, Washington State Univ., Pullman, WA, 99164
* Corresponding author (j-dixon{at}tamu.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: Db, bulk density • FH, ferrihydrite • LBNL, Lawrence Berkeley National Laboratory • PAM, polyacrylamide • SEM, scanning electron microscope • TEM, transmission electron microscope • XRD, X-ray diffraction
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This paper approaches soil structure development from the constructive side by assembling soil constituents and measuring their influence on two important properties: bulk density (Db) and hydraulic conductivity. Investigating avenues for improving soil structure is a timely pursuit. There are extensive areas of soils low in organic matter and large quantities of mine spoils (Senkayi et al., 1983) where development of soil structure would benefit long-term productivity. At the same time, there are abundant quantities of soil aggregating agents such as organic municipal wastes (Clapp et al., 1994) and wastes from confined feeding of farm animals.
By employing only the clay fraction (i.e., excluding silt and sand), information can be gained on the influence of aggregating agents (Fe oxide and organic substances) in formulated soils. Ferrihydrite (FH) was used as a representative of the Fe oxides because it is the most reactive Fe oxide. It has a very high surface area, a positive charge at most soil pHs (Bartoli et al., 1992), a low bulk density, and abundant water retention (2.70 kg kg–1) in natural occurrences (Wells and Childs, 1988). Ferrihydrite has been reported in many soil environments and seems to be a precursor for the formation of hematite (Allen and Hajek, 1989; Schwertmann and Taylor, 1989). Also, FH can be managed in the field by changing cultural practices and controlling seasonal water content (e.g., paddy versus dryland crops) (Wang et al., 1993). Ferrihydrite is the first Fe oxide to form under some conditions and therefore precipitates in soils undergoing sudden changes in redox potential. Polyacrylamides (PAMs) were utilized as surrogates for soil organic matter because they can improve soil aggregation (Sojka et al., 1998) and are available in different ionic forms.
The presence of halloysite in soils impacts soil properties and behavior because the particles usually are cylinders, spheres, or cup-shaped and cannot be packed as densely as minerals such as platy kaolinite. Many strongly weathered soils (e.g., Ultisols) from tropical regions have bulk densities ranging from 0.6 to 1.0 g cm–3, whereas soils from temperate climates have bulk densities from 1.2 to 2.0 g cm–3 (Wesch, 1999; Brady, 1990). Bulk densities of halloysite clay mixed with foundry molding sands at various water contents were lower than such mixtures containing kaolinite (Grim, 1962). Low bulk densities are associated with high porosity.
Movement of water through natural soil is complex because of the wide range of pore sizes and the multiple sampling and statistical methods required to determine patterns of water movement (Schwartz et al., 1999). In this study, we chose to use homogenous starting materials of clay size to investigate the hydraulic conductivity of the finer pores in fabricated soils. Water movement in soils is most rapid where the tortuousity is least, assuming equal pore size. Small halloysite particles produce a less tortuous flow path than the wider kaolinite flakes that are prone to be oriented horizontally. As noted earlier, smooth, platy particles of kaolinite clay are expected to settle into a higher Db mass than do tubular particles of halloysite that have a calculated Db of 90.7% of the intrinsic density. In addition, Fe oxides are hypothesized to promote porous aggregates because of particle roughness, water content, and inter-particle bonding. As a result, tubular clay particles and Fe oxides increase the porosity, and therefore, the permeability of the soils. Organic matter contributes to aggregation in soils (Brady and Weil, 1999). Thus, the objectives of this research were to quantify the influence of clay morphology, FH, and PAMs on Db and/or hydraulic conductivity in fabricated clay soils.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Preparation of Formulated Soils
Kaolinite and halloysite of clay-size, <2µm were separated from the sand and silt fractions where quartz and other contaminants tend to be concentrated. Ten grams of each whole mineral sample was placed into a 250-mL centrifuge bottle, 200 mL of 0.05 M sodium chloride was added, and the sample then was agitated on a mechanical shaker for at least 24 h. After the sample was removed from the shaker, it was centrifuged at 283 x g (1000 rpm) for 5 min, and the clear supernatant was decanted off. To remove the sand, 200 mL of pH 10 Na2CO3 was added to the bottle, the sample was resuspended by manual shaking, and then allowed to settle for a few minutes. The suspended material (silt and clay) was then poured through a 45 µm (325-mesh) sieve into a 4-L beaker. The addition of pH 10 Na2CO3 was repeated until the supernatant was clear. An ultrasonic probe was utilized to disperse the particles during each washing cycle.
The suspension that passed through the sieve was then poured into 250-mL centrifuge bottles and centrifuged at 159 x g (750 rpm) for 3.2 min according to Stokes law. The supernatant, which contained the clay fraction, was decanted into a 4-L beaker. This process was repeated until all of the clay/silt mixture was separated, and then the samples were washed several times with pH 10 Na2CO3 until clay no longer appeared in the supernatant.
To flocculate the clay fraction, a minimum amount of 0.5 M MgCl2 was added to the beakers to obtain visual flocculation (Kunze and Dixon, 1986). After the clay had settled out of the solution, the clear supernatant was decanted; the remaining clay suspension was poured into 250-mL centrifuge bottles, and centrifuged at 283 x g (1000 rpm) for 5 min. Then the samples were rinsed several times with distilled water to remove the MgCl2. The halloysite samples were rinsed eight times and the kaolinite samples were rinsed three times until the clay began to disperse.
The Fe oxide, FH, was synthesized from potassium hydroxide and ferric nitrate and was estimated to have a surface area of 200 to 300 m2 g–1 (Schwertmann and Cornell, 1991).
Three commercial PAM surfactants were used to simulate soil organic matter. The surfactants used included one cationic (Magnifloc 494C), one anionic (Magnifloc 836A), and one uncharged (Magnifloc 903N) from Cytec Industries, West Patterson, NJ. Three concentrations of each surfactant solution were made (2.5, 0.25, and 0.025 g L–1).
Bulk Density
For the Db experiments, 1 g of clay was placed in a 250-mL centrifuge bottle, and FH was added. The amount of FH added constituted 0, 1, 10, 20, 25, and 30% of the total mixture. Two replicates were made for both halloysite and kaolinite for the 0, 1, and 10% concentrations. However, only one sample for each clay mineral was made with 20, 25, and 30% concentrations of FH.
The pH of each sample was checked; if the pH was below 6.0, 1% (v/v) sodium carbonate was added to raise the pH to approximately 7. These mixtures were brought to 100 mL total volume by adding distilled water, homogenized by shaking, and then transferred to 40-mL centrifuge tubes. They were then centrifuged for 10-min intervals at different centrifugation speeds (11 to 5519 x g, 200 to 4400 rpm) until they obtained a constant volume of clay (Jackson, 1969). After each centrifugation, the compaction depth of the clays was measured and recorded. These clay depths and the tube diameters were used to calculate Db.
 |
Two replicates were made for both halloysite and kaolinite with each concentration of organic surfactant. The same method used with the FH experiments was used for these mixtures, except the pH was not checked and brought up to 7.0. After the experiments were completed, the pH of the samples was determined. Values ranged from 5.5 to 5.8 for halloysite samples, and from 4.7 to 6.5 for the kaolinite samples.
Hydraulic Conductivity
Hydraulic conductivity was measured for halloysite and kaolinite samples with 0, 10, and 30% FH (four replicates for each mixture) using the falling-head method with a novel dual purpose centrifuge tube and leaching column constructed from Lexan plastic tubing closed with a snap-cap (Kimble, Vineland, NJ augmented with Parafilm, Chicago, IL [Fig. 1b]). In this case a circular piece of Fisher P5 filter paper (Fisher Scientific, Pittsburgh, PA) was added to the bottom of the tube and the closure was replaced (Fig. 1b). The tube was protected from deformation from the force of gravity by surrounding it with water in each centrifuge trunnion cup. Approximately 2 g of the dispersed sample was poured into each tube, and then centrifuged three times at 103 x g for 30 min to form a solid plug. Once the plug height reached equilibrium, the cap was carefully removed and cheesecloth was secured over the end to hold the filter paper in place.
|
 |
After conducting the experiment, the caps were replaced on the columns, and the samples were centrifuged three times at 1029 x g for 30 min until an equilibrium height had been reached; hydraulic conductivity data were collected again. This same procedure was repeated to collect data for the samples run at 1642 x g. After all desired data were collected, the samples were placed in a 60°C oven. Once dry, the mass of each sample was measured, and the data were used to calculate bulk density. Additionally, pH of the samples was measured by rehydrating the samples with a 3:1 water/sample ratio. Values ranged from 4.4 to 6.5 for the kaolinite samples, and from 4.5 to 5.6 for the halloysite samples (samples with more FH had lower pHs).
Electron Microscopy
Scanning electron microscopy (SEM) was performed on air-dried material from the centrifuge tubes after the hydraulic conductivity was determined. Broken surfaces were examined on a JEOL 6400 (JEOL [U.S.A.] Inc., Peabody, MA) after adhesion to an Al sample stub using conductive carbon paint. The SEM micrographs were obtained at 20 KeV at a working distance of 12 mm.
The transmission electron microscopy (TEM) was performed on dispersed clay samples taken before compaction by centrifugation. The samples were drop mounted on Type A silicon monoxide films on copper grids (Ted Pella Inc., Redding, CA) except for the sample shown in Fig. 5a that was on a holey C film on a Cu grid. All TEM examinations were performed with a JEOL 2010 microscope (JEOL [U.S.A.] Inc., Peabody, MA) at 200 KeV.
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Particle Morphology Shown by Scanning Electron Micrographs
The platy kaolinite particles are visible in the scanning electron micrographs of samples that had been compacted by centrifugation and dried (Fig. 3a). Many sizes of particles are visible and the thicker particles are prominent in edge view in the bottom center of the field. Also, many small plates are incongruously situated on the edges of the larger and thicker ones. The particles appear to be randomly oriented and some large pores are visible. The many thin particles about a micrometer across are easily overlooked on the thick stacks. The many stacks of kaolinite plates raise the question of their reassembly during compaction.
|
Particle Morphology Shown by Transmission Electron Micrographs
The transmission electron micrographs of both minerals show how smooth and well dispersed the clay particles were (Fig. 4a and 5a). The contrasting morphology of the two minerals limits geometric comparison. Kaolinite plates are thin and the clean abrupt edges indicate the lack of dense particles of Fe oxide (Fig. 4a). Also, they are euhedral six-sided plates that exhibit uniform absorbance of the electron beam indicating high sample purity. Halloysite, on the other hand, is generally tubular, but the tubes vary in length, thickness, and shape of the terminations. Kaolinite and halloysite samples containing added FH (Fig. 4 and 5) had many dark granular particles of FH that were precipitated in solution and added as colloids that had never been dried. The FH is more opaque than the layer silicates because of its higher absorbance in the electron beam. The interstitial space is largely free of FH in samples of both minerals suggesting association of the oxide particles with the layer silicates before drying from water suspension on the support membrane.
|
Halloysite–Ferrihydrite Association
Halloysite particles are smooth and free of granular attachments (Fig. 5a). The inset at 1 is a segment of the particle below it that demonstrates how smooth the particles are even when enlarged. One notable exception is the tube with notches below and to the right of the inset. It appears to have formed adjacent to other tubes that prevented its full growth and left the semicircular notches. The hole in the carbon support membrane enhances the view of interior layer separations parallel to the tube axis and shows the smoothness of the exterior of the two particles that overlay the hole. Adjacent to Particle 2 is a wide assortment of tube lengths illustrating the diversity in aspect ratio in this halloysite sample. Unfortunately this holey carbon grid support membrane is not as clean as the silicon monoxide membranes used for the other micrographs and two small dark contaminant spots are evident adjacent to the Particle 2 and a few other lighter ones are visible near the bottom of the micrograph.
An overview of the halloysite micrographs (Fig. 5b–d) shows that very little FH is independently resting on the support membrane suggesting that the oxide was associated with the layer silicate tubes before the suspension was dried on the support membrane. The inter-particle space is clean indicating complete particle association between FH and halloysite.
The presence of 10% FH added to the halloysite is evident as irregular aggregates in the micrograph at 1, 2, 3, and 4 (Fig. 5b). Also, FH is present as many small particles attached to the halloysite tubes that are most evident in the enlarged inset at 5 from the adjacent particle to its left. Numerous particles of FH are evident in this micrograph of the sample containing 20% FH (Fig. 5c). The inset particle segment at 3 shows the granular particles of FH throughout neglecting the small plate, presumably kaolinite, near the top of the inset cut from the tube to its left. This sample (Fig. 5d) contains 30% FH and the halloysite particles consistently have associated small or large particles of FH that give the tubes an overall appearance of rough surfaces. The FH particles are profuse from small to large on the halloysite tube segment in enlarged Inset 1. Ferrihydrite is porous and highly hydrated when formed, thus it does not absorb the electron beam as completely as solid Fe oxides like goethite.
The distribution of FH at the three treatment levels in halloysite contrasts with the kaolinite treatment results discussed above. As the amount of FH was increased the numerous small particles that adhere to the tubes became more prominent. The frequency of FH particles on the halloysite tubes is much greater than on the kaolinite plates and the size range of the FH particles is smaller on the halloysite.
Bulk Density versus Centrifugal Force
Increasing centrifugation speed in steps to obtain the calculated force from 11 to 5519 x g increased sample bulk density of both sets of formulated soils as expected (Fig. 6). The addition of FH decreased the Db of both clays yet the amount of FH added appeared unimportant for kaolinite (Fig. 6). In contrast, halloysite bulk densities were progressively decreased by each additional amount of FH including the 30% level. The smaller stepwise additions of FH illustrated in Fig. 7 indicate that for kaolinite 10% FH leveled off and Db was uniform to 30%. Halloysite minimum Db apparently was reached at about 20% FH, which is consistent with the higher surface area of halloysite than kaolinite.
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Halloysite–Ferrihydrite Association
There is growing evidence that halloysite has isomorphous substitution creating limited negative charge sites on the particles (Bailey, 1989; Tari et al., 1999). The presence of limited substitution favors sorption of FH that is assumed to have many positive sites since it has a zero point of charge between pH 7 and 9 (Bartoli et al., 1992; Bigham et al., 2002). The pH range of our samples is below neutrality. Also, a special affinity of goethite for halloysite was reported for a New Zealand halloysite (Golden and Dixon, 1985). The halloysite was mostly removed with the goethite by magnetic fractionation, indicating a strong Fe oxide–halloysite association.
Kaolinite and Ferrihydrite Relationships
Kaolinite and FH mixtures compacted by centrifugation formed porous tabular masses of FH only loosely associated with the kaolinite plates in contrast to the frequent association of irregular particles of FH that adhered to similarly compacted mixtures of halloysite and FH. The kaolinite plates apparently pressed the FH into sheet-like masses and the absence of charge on the kaolinite plates did not produce compound particles as for halloysite and FH. This tendency for FH to be shaped by pressure as a colloid may contribute to its occurrence in Fe concentrations in soils and sediments, for example, placic soil horizons and bog Fe ores (Schwertmann and Taylor, 1989).
Ferrihydrite Colloids
Ferrihydrite is one of the few soil minerals that is gelatinous and has been observed to form and dissolve seasonally (Wang et al., 1993). It occurs only as small disorderly crystals well within the colloidal range when fully dispersed. Also, it has spheroidal particles approximately 5 nm in diameter (Schwertmann and Taylor, 1989). Recent investigations have shown that the number of colloidal spheres (10–7–10–9 m) in each packing polyhedron and density range is the same (Manoharan et al., 2003). Considering FH as an assembly of (hollow) colloidal spheres offers a new avenue for investigating this poorly crystalline soil mineral.
Applications
A thorough evaluation of the range of adsorptive versus structural properties of kaolinites and halloysites is needed to better define the characteristics of these two minerals. Abnormally high cation-exchange capacities for kaolinite in geologic specimens was attributed to smectite based on lattice fringes intimately associated with the kaolinite. Yet the smectite was not detected by XRD (Ma and Eggleton, 1999). Our observations in this paper indicate the need for investigations of the composition of halloysite and kaolinite in soils at the nanometer scale. The breadth of the definition of halloysite is called into question by the results of this study and related references on its charge properties.
Halloysite frequently forms under wet conditions as indicated by the locations where it occurs and the presence of interlayer water in the 1-nm phase. It forms in the rock–soil interface zone (Calvert et al., 1980), in hydrothermal springs and subtropical weathering of rhyolite volcanics (Harvey and Murray, 1993), in cracks in tonstein (Senkayi et al., 1984) and around weathering of granite boulders (observed in Alabama by J.B. Dixon). These observations suggest that the presence of halloysite in soils and saprolite (our sample) is associated with movement of surface water to ground water. Thus, understanding the hydraulic behavior of halloysite is useful for predicting movement and properties of ground water. A better understanding of the sorption properties of halloysite for FH and surfactants should help in understanding the chemistry of ground water.
Soils containing halloysite appear to have more potential for structural improvement by FH aggregation than soils containing the same amount of kaolinite. Halloysite has been reported in soils at 45 locations throughout the world (Allen and Hajek, 1989). It is most often reported in volcanic soils that abound in Central America, Japan, New Zealand, and in the USA in Oregon and Hawaii. Halloysite also has been reported in soils of Georgia, North Carolina, Texas, and Alabama. In view of the weak XRD curve of halloysite probably it is sometimes overlooked or confused with kaolinite, which is ubiquitous in soil clays.
Partial success of PAM organic polymers as a surrogate for soil organic matter in fabricated soils clays deserves further evaluation. This paper is a small step toward forming a synthetic soil. It is worthy of note that the Lawrence Berkeley National Laboratory (LBNL) in California established the first synthetic biology department in 2003 (Ferber, 2004). The LBNL example provides a challenge to soil scientists to fabricate synthetic soils for research and instruction to sharpen concepts and quantify biological, chemical, mineralogical, and physical reactions in soils.
|
ACKNOWLEDGMENTS |
|---|
Received for publication August 11, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  W. Markgraf and R. Horn Scanning Electron Microscopy-Energy Dispersive Scan Analyses and Rheological Investigations of South-Brazilian Soils Soil Sci. Soc. Am. J., May 16, 2007; 71(3): 851 - 859. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. JOUSSEIN, S. PETIT, J. CHURCHMAN, B. THENG, D. RIGHI, and B. DELVAUX Halloysite clay minerals -- a review Clay Minerals, December 1, 2005; 40(4): 383 - 426. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
||||
