SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 16 May 2007
Published in Soil Sci Soc Am J 71:851-859 (2007)
DOI: 10.2136/sssaj2006.0231
© 2007 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
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 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 Google Scholar
Google Scholar
Right arrow Articles by Markgraf, W.
Right arrow Articles by Horn, R.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Markgraf, W.
Right arrow Articles by Horn, R.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Markgraf, W.
Right arrow Articles by Horn, R.
Related Collections
Right arrow Soil Physics

SOIL PHYSICS

Scanning Electron Microscopy–Energy Dispersive Scan Analyses and Rheological Investigations of South-Brazilian Soils

Wibke Markgraf* and Rainer Horn

Institute for Plant Nutrition and Soil Science, Hermann-Rodewald-St. 2, Christian-Albrechts-Univ. to Kiel, 24118 Kiel, Germany

* Corresponding author (w.markgraf{at}soils.uni-kiel.de).


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Scanning electron microscopy, including energy dispersive scan analyses and amplitude sweep tests with controlled shear deformation, were conducted on four kaolinitic, Fe-oxide-rich Brazilian Hapludox soils and one montmorillonitic Calciudert soil to elucidate results obtained from a parallel-plate rheometer. In the Brazilian Hapludox soils, the occurrence of pseudosand (an aggregation of fine particles with a grain size of sand), due to Fe (hydr)oxide cementation, had a significant effect on aggregation and microstructural stability. We could observe the presence of hematite, goethite, and, in rare cases, halloysite, as the major stabilizing minerals in such soils. The influences of soil organic matter (SOM), Fe oxides, and clay minerals on micromechanical shear behavior under oscillation were tested under saturated and predrained (at –60 hPa) conditions. Collected data included G (shear modulus), G' (storage modulus), G'' (loss modulus), linear viscoelastic deformation range, deformation limit ({gamma}L), and yield stress ({tau}y). From the data, the dissipation of elasticity was derived, and the values ranged from 200 to 1000 Pa in all untreated samples, both Hapludox and Calciudert soils, and to 50 to 500 Pa in SOM-leached Hapludox samples. Minimum {tau}y values of 10 to 50 Pa occurred after a sodium dithionite treatment (Fed leached) of the four Hapludox samples. These findings show that Fe (hydr)oxides have a more stabilizing effect on the microstructural stability than SOM.

Abbreviations: CSD, controlled shear deformation • EDS, energy dispersive scan • Fed, sodium-dithionite-soluble Fe • {gamma}L, deformation limit • G', storage modulus • G'', loss modulus • HAC, high active clay • LAC, low active clay • LVE, linear viscoelastic • SEM, scanning electron microscopy • SOM, soil organic matter • {tau}y, yield stress • {omega}, angular frequency


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soil mechanics is an intertwined combination of solid and fluid mechanics. In soil mechanics studies, both the stiffness and strength of the skeleton, and the permeability and porosity of the pore space, are highly significant variables (Bolton, 2000). The type of soil, including its physicochemical compounds (i.e., clay minerals), and the shear stress to which a soil is exposed are of great importance.

General stress–strain relations for soil mechanics have been described by Mitchell and Soga (2005), who use terms such as the shear modulus (G, MPa), shear strain ({gamma}, %), and the strain rate ({varepsilon}, %), which are dependent on time or frequency. Besides well-established methods such as direct-shear or triaxial tests, rheometry appears to be another suitable method, whenever mechanical behavior between single particles (tactiles, platelets, and grains) needs to be measured. These measurements can include quantifying effects due to both soil physical and chemical factors or to preconditions that affect both of them.

Based on fundamental knowledge of soil mechanics, stress–strain relations with small strains can be transferred to rheological tests by using amplitude sweep tests under oscillatory conditions. Amplitude sweep tests include information about stress–strain relations, namely the relation of shear moduli G' (storage modulus) and G'' (loss modulus) to a controlled shear deformation (CSD) {gamma}. In amplitude sweep tests with CSD, the shear strain, as input, is set as a logarithmic ramp, and a constant frequency f or angular frequency {omega} is provided. Furthermore, calculations of a deformation limit {gamma}L—related to strain—or a yield stress {tau}y—related to stress—determined from a collected database, lead to conclusions about the stress–strain relation on a microscale. In addition, graphs of G' and G'' give specific characteristics of the material under study. The shear behavior is influenced by texture (clay content, clay mineralogy, Fe [hydr]oxides). Physicochemical properties and water content affect microstructural changes, (micro)aggregation, and microstructural stability. By interpreting the curve characteristics of G' and G'', the following can be quantified: dissipating elasticity (= decrease of G') and frictional heat (= slight increase of G'' before transition to creeping, which is related to an increase in contact points and structural breakdown). Rheology often has been applied to study the impact of (cat)ions on structureless clay pastes in concrete, where a defined structure exists between cement and larger particles (Szecsy, 1997; Akroyd and Nguyen, 2003; Phan and Chaouche, 2005). Rheology has rarely been applied to soils to study the internal microstructure, however, particularly under arable conditions, where land management will have an impact (Horn and Baumgartl, 1999).

Some suitable approaches have been developed in other fields such as clay science and material science (Jasmund, and Lagaly, 1993; Schulz, 1998; Mezger, 2002). Rheology also has not been related to descriptions of soil aggregation. The mechanisms that lead to a stabilization of soil by SOM and Fe (hydr)oxides and the role of clay might be quantified by this method.

Markgraf et al. (2006) and Markgraf and Horn (2006a, 2006b) investigated the influence of texture, natural physicochemical properties, water content, and valence (hydration mechanisms) on the micromechanical shear behavior of soil. They based their studies on the work of Keedwell (1984), Vyalov (1986), and Ghezzehei and Or (2001). They found that coarse, sandy substrates show a turbulent shear behavior compared with clay-rich materials, which show a sliding shear behavior. They defined clay mineralogical compounds like kaolinite as low-active clay (LAC) and smectites as high-active clay (HAC) minerals. They found that lower water contents (in samples predrained at –60 hPa) led to a higher structural stability due to menisci forces. With regard to hydration mechanisms, covalent bonds (Ca2+, Mg2+) are stronger than monovalent ionic forces (Na+) (Markgraf and Horn, 2006a, 2006b). Markgraf and Horn (2006b) described results from scanning electron microscopy (SEM) studies and concluded that additional information about surface properties and the shape of (micro)pores and cracks needs to be obtained to explain the rheological phenomena.

Application of SEM in clay mineralogy is common. For example, Abichou et al. (2002) investigated microstructural changes and hydraulic conductivity of simulated sand–bentonite mixtures. They applied two methods: photomicrographs using digital image acquisition software, and scanning electron micrography. The imagery helped to explain the importance of soil microstructure and resulting differences in curves of G' and G''. Differences in the relation of elastic and viscous parts of the curves, and in values of deformation limits {gamma}L (with a specific {tau}y in amplitude sweep tests), also can be determined from SEM studies. These visual findings may also lead to conclusions about micromechanical characteristics on a particle-to-particle scale because they give information about particle shape and size, which influence shear behavior. Coupled mechanical and SEM measurements in soils are rare, although they can provide insight into geomechanical processes.

Soil organic matter and Fe (hydr)oxides are well known to influence the structural stability and aggregation of soils across a wide range of types. Differences in the mechanical behavior of soil are also observed at larger scales; however, only a few studies of soil-strength changes as related to Fe (hydr)oxide and SOM can be found. Muggler et al. (1999) quantified differences in the grain-size distribution determined by laser diffraction after three pretreatments. They followed up studies of Deshpande et al. (1968), Tisdall and Oades (1982), Oades and Waters (1991), and Duiker et al. (2003). Duiker et al. (2003) shook samples with water, leaching them of organic matter, and then followed that treatment with deferration. Muggler et al. (1999), who studied Brazilian Oxisols from Minas Gerais, demonstrated well-defined differences in aggregation, depending on the parent material (either rock saprolites or sediment) and the formation or remobilization of Fe. The mechanical strength was not determined, however, although it was important in understanding the structure-formation processes and the corresponding effects on soil strength. This information is also important to understand the interaction between the physical and chemical processes that strengthen particles.

Rheological investigations of bentonite suspensions (Na-bentonite Ibeco Seal-80 in Markgraf et al., 2006) and silt- and clay-rich substrates corroborate the hypothesis that the application of a parallel-plate rheometer (MCR 300, Anton Paar Co., Stuttgart, Germany) is appropriate for the quantification of microstructural changes on a particle scale (Markgraf et al., 2006).

The parallel-plate rheometer provides information about aggregate stability (Breuer and Schwertmann, 1999) and pseudosand effects, as defined by Cornell and Schwertmann (2003). Duiker et al. (2003) used the method to demonstrate the effects of Fe (hydr)oxide crystallinity on soil aggregation. By executing oscillatory tests (amplitude sweep tests) with CSD, deformation effects due to vibrations, which are produced by farm implements, can be simulated (Horn and Baumgartl, 1999; Garciano et al., 2001). Such measurements have not been performed, however, including quantification of the strength of the various bonding mechanisms.

This study was done to prove or disprove the following hypotheses:

  1. Soil organic matter, as well as Fe (hydr)oxides, increase the mechanical soil strength and result in the occurrence of pseudosand.
  2. Curve characteristics, as well as single parameters like G', G'', the calculated linear viscoelastic (LVE) deformation range, and {gamma}L, enable a quantification of dissipating elasticity (loss of rigidity) in soils on a microscale (particle–particle scale). As a result, amplitude sweep tests will show significant differences in these characteristics in Hapludox and Calciudert soils.
  3. The SEM–energy dispersive scan (EDS) analyses of untreated, H2O2–treated, and sodium-dithionite-treated samples facilitate the explanation of microstructural properties and changes. Results depend on these preceding treatments and affect the rheological properties.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Measurements were performed on two types of soil: four kaolinitic Typic Hapludox soils (Soil Survey Staff, 2003; Latossolos Vermelho according to da Silva et al. [2004] and Streck et al. [2002]), which contained various compounds of hematite and goethite and different amounts of SOM and clay, and a montmorillonitic Typic Calciudert soil (Soil Survey Staff, 2003; Vertissolo Ebânico according to Streck et al. [2002]).

Geography and Geology
The southernmost state of Brazil, Rio Grande do Sul, is climatologically classified as humid subtropical (Cfa climate according to Köppen, 1931). Temperate, natural grasslands cover the southern part of Rio Grande, and temperate Brazilian pine [Araucaria angustifolia (Bertol.) Kuntze] forests predominate in the northern, higher regions (planalto). These two major vegetation units have geomorphologically separate structural elements. They are divided by an escarpment into the upper planalto or Serra Geral and the lowlands, which include the depressão periférica and the escudo. These both belong to the pampa region. The Paraná basin was formed on the Brazilian shield, in which marine and continental sediments have been deposited. At the southern border of this formation (Santo Ângelo 28°12' S, 54°13' W, região das Missões, and Cruz Alta 28°34' S, 53°06' W, região do Planalto Medio; da Azevedo, 2003), sediments are covered by effusive basic rocks (Cretaceous basalt) of the Serra Geral. Santana do Livramento (31°06' S, 55°12' W), which is located at the Uruguayan border, has sedimentary rocks and fluvial Mesozoic Botocatú sandstones of the Brazilian shield as the parent material. Both Serra Geral and Botocatú belong to the São Bento formation (Pinto, 1966; FAO, 1971; Silvério da Silva and Menegotto, 2002; West et al., 2004).

Substrates
Samples were taken from three locations: ≥40-cm depth under no-tillage (at least 5 yr) and campo-natural conditions (meadow or natural forest) in Santo Ângelo (Typic Hapludox, kaolinitic), in Cruz Alta (sandy and clayey Typic Hapludox, kaolinitic), and in Santana do Livramento under pasture (Typic Calciudert, smectitic). Homogenized air-dried samples were sieved to <2 mm and were repacked in 45-cm3 cylinders (n = 3, bulk density of air-dried 2-mm sieved soil = 1.4 g cm–3). Thereafter they were completely saturated with distilled water; parallel samples (n = 3) were prepared and drained at –60 hPa (Table 1). Altogether, 78 samples were prepared: 30 of untreated, natural soil material from Santana do Livramento, Cruz Alta, and Santo Ângelo; 24 of H2O2–treated; and 24 of Fed–leached samples (Cruz Alta and Santo Ângelo).


View this table:
[in this window]
[in a new window]

  		Table 1. Physical and chemical properties of investigated samples from Cruz Alta (sandy and clayey Typic Hapludox), Santo Ângelo (Typic Hapludox under natural forest [F] and no-tillage [NT]) and Santana do Livramento (Typic Calciudert).

 
Analyses
Analyses were conducted according to standard methods as described in Schlichting et al. (1995), and van Reeuwijk (2002). Sieved (<2-mm) and homogenized samples were taken to measure exchangeable cations, which were extracted by 1 M NH4OAc. Concentrations of Ca2+ and Mg2+ were measured by an atomic absorption spectrometer, whereas K+ and Na+ were measured by flame emission. Iron oxides were extracted by sodium dithionite according to Mehra and Jackson (1960). Soil organic matter was removed with H2O2.

Typic Hapludox soils are dominated by a kaolinitic clay fraction, while Typic Calciudert soils have smectitic clay mineralogy and a typically high content of Mg2+ and Ca2+ in a ratio of 1:2.5. Sandy and clayey Typic Hapludox soils from Cruz Alta showed textural differences, whereas clayey Typic Hapludox soils from Santo Ângelo contained variable amounts of soil organic matter (here total C) contents. X-ray diffractometry was performed by Prof. Dr. R. Jahn, mLU Halle, and was conducted according to standardized methods as described in Whittig and Allardice (1986).

Amplitude Sweep Tests
Markgraf et al. (2006) gave a detailed description of amplitude sweep tests, including a theoretical background to rheology and its application. A parallel-plate rheometer, MCR 300, was used. During all tests a constant temperature of 20°C was maintained, regulated by a Peltier unit.

Amplitude sweep tests under oscillatory conditions were conducted, with CSD {gamma} = 0.0001 to 100%, {omega} = {pi} s–1 (f = 0.5 Hz), and 30 measuring points, which led to an average test duration of 16 min. A plate distance of 4 mm was preset according to a plate radius of the rotating bob of 25 mm and the given texture (>2 µm). The tests were controlled by the software US 200 (Anton Paar Co.).

A representative result of an amplitude sweep test is shown in Fig. 1. The plots of G' and G'' are generated automatically during a test. Three phases of elasticity loss can be identified as follows:

Phase 1. Initial or plateau phase, G' > G''; an elastic behavior is observed, represented by a spring for ideal elastic substances according to Hooke's law. An LVE range and the included {gamma}L are parameters needed to quantify "stored elasticity" of any viscoelastic substrate, e.g., soils.
Phase 2: Stage of transgression or the intersection of G' and G.''
Phase 3: Final stage of structural collapse, G' < G''; a viscous character predominates—substances are creeping or running. This behavior can be represented by a dashpot (a device for damping shock or vibration), an analog for ideal fluids according to Newton's law.


Figure 1
View larger version (49K):
[in this window]
[in a new window]

  		Fig. 1. Plots of G' (storage modulus) and G'' (loss modulus). Three stages of elasticity loss can be defined, showing a gradual transition of an elastic (G' > G'') to a viscous (G' < G'') character.

 
More parameters need to be identified for curve progressions: (i) the duration of each phase of elasticity dissipation; (ii) characteristics of the plateau phase, distances between graphs of G' and G'', slope progression in Phase 2, the intersection of G' and G''; and (iii) whether the state G' < G'' is reached or not in Phase 3, which is the stage of microstructural collapse where the final level of G' and G'' is considered.

For calculations of the LVE deformation range, {gamma}L, and {tau}y (that is, the Yield Stress II), analyses were executed after each completed test run (Ghezzehei and Or, 2001; Mezger, 2002; Markgraf et al., 2006; Markgraf and Horn, 2006a, 2006b). To do this, under oscillatory conditions a tangent is fit to the G' curve, which is based on the minimum {gamma} value and is limited by a decline of G' that has a deviation of >5% in relation to this calculated tangent.

Scanning Electron Microscopy
Scanning electron microscopy was done with a CamScan CS 44 (E.O. Electron-Optik Service GmbH, Dortmund, Germany), which also can be used for EDS investigations. Oven-dried (at 40°C) samples and Al holders are connected with a self-adhesive C die. For SEM, required conductivity was achieved by applying a gold-palladium coating under high-vacuum conditions (sputtering). The SEM micrographs were obtained at 15 keV at a working distance of 15 mm. Monochrome photographs were taken with a small picture reflex camera, which is integrated into the CamScan CS 44 working station as an external unit. A detailed description of SEM and microanalysis (and their applications) is given by Henning and Störr (1986, see their tables) and Schmidt (1994). Visual images of polarizing microscopy were prepared according to the method of van Reeuwijk (2002).

Water Content
The water content (w/w) was determined before rheological tests according to standardized methods (Hartge and Horn, 1999).

Statistics
Seventy-eight samples were tested, including two repetitions (n = 3). Seven to eight samples with different treatments—either saturated untreated, Fed– and SOM-leached, or predrained at –60 hPa—were measured per day, which resulted in 21 to 24 measurements per day, with 14 to 18 min for each test. From data of {gamma}L and {tau}y, arithmetic mean values were calculated based on a preset range (in US 200) of tolerance with ±5% deviation included in Yield Stress II analyses. Hence, a high level of accuracy can be assumed. This is also shown by the fact that a minimum deformation of 0.0001% is equivalent to the deflection of 1 µm under oscillatory conditions.


   RESULTS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Amplitude Sweep Tests
Microstructural effects of SOM and Fed, which depended on the water content, were obvious. Values of {gamma}L and {tau}y showed significant differences (Table 2). Generally, under predrained conditions, {gamma}L and {tau}y increased, except in the untreated clayey Typic Hapludox samples from Santo Ângelo. Furthermore, a decrease of {gamma}L and {tau}y becomes obvious if untreated samples are compared with SOM- and Fed–leached samples.


View this table:
[in this window]
[in a new window]

  		Table 2. Summarized results from amplitude sweep tests (with controlled shear deformation). Values of the deformation limit ({gamma}L) and yield stress ({tau}y) are arithmetic means, n = 3.

 
Untreated samples, including all natural compounds, were more stable than SOM- or Fed–leached samples. In the latter case, both {gamma}L and {tau}y decreased noticeably, which can be summed up as follows: untreated > SOM leached > Fed leached (Fig. 2b–2e). In addition, in almost every case, {gamma}L and {tau}y increased under predrained conditions (–60 hPa; Fig. 2a and 3a–3d, Table 2). A secondary stabilizing effect was demonstrated: levels of G' –60hPa and G'' –60hPa plots were higher those for saturated conditions. Increases of G' and G'' showed a higher rigidity, assuming that G' is higher than G'' in the first phase.


Figure 2
View larger version (42K):
[in this window]
[in a new window]

  		Fig. 2. Results of amplitude sweep tests with (distilled water) saturated samples of (a) a Calciudert (and under predrained conditions), Santana do Livramento, (b) a sandy and (c) a clayey Hapludox from Cruz Alta, (d) a clayey Hapludox under natural forest, and (e) a clayey Hapludox under no-tillage conditions, both from Santo Ângelo. G' = storage modulus, G'' = loss modulus, {gamma} = shear strain, SOM = soil organic matter, Fed = dithionite-soluble iron.

 

Figure 3
View larger version (65K):
[in this window]
[in a new window]

  		Fig. 3. Results of amplitude sweep tests with predrained (at –60 hPa) samples of (a) a sandy and (b) a clayey Typic Hapludox from Cruz Alta, (c) a clayey Typic Hapludox, Santo Ângelo, under natural forest, and (d) a clayey Typic Hapludox, Santo Ângelo, under no-tillage conditions. G' = storage modulus, G'' = loss modulus, {gamma} = shear strain, SOM = soil organic matter, Fed = dithionite-soluble iron.

 
An ideal behavior is demonstrated with the data on the untreated saturated and predrained Calciudert samples (Fig. 3a). A sliding shear behavior predominates, which results from a smectitic clay mineralogy. The plateau phase, which is characterized by a parallel run of G' and G'', is well defined, followed by a typical progression in Phase 2. It ends with the intersection of G' and G''—at a lower deformation input under predrained conditions—and finally leads into Phase 3, the stage of viscoplastic behavior.

When patterns of the variously treated Hapludox samples are compared with the untreated Calciudert samples, characteristics regarding textural compounds, properties of microcracks, cavities, and pores, and clay mineralogy (HAC and LAC) are evident. Single grains, kaolinite piles, and small roots are embedded as an assemblage with an Fe-oxide coat containing fissures (Ohtsubo et al., 1991). These fissures are preferentially at connection points under untreated conditions. Organic matter was completely removed in H2O2–treated samples. It can be assumed that binding mechanisms, which may have been affected by, e.g., (micro)roots and their exudates, were reduced or disabled. After the sodium dithionite treatment, surfaces of single grains became bare. Kaolinite piles may function as single grains with regard to shear behavior, if one assumes stable structural conditions of partially sharp-edged grains. In this case, a direct surface-to-surface or edge-to-edge contact can be assumed during amplitude sweep tests, which leads to a higher angle of friction. This is indicated by a lower level of G' and G'' and a decelerating set in Phase 3. Based on such visual findings, a link to rheological parameters can be done, assuming that single grains show a different mechanical behavior than more coherent structures such as microaggregates. When combined with SEM–EDS findings, characteristics of G' and G'' lead to an understanding of friction processes on a particle-to-particle scale.

Phase 1 is less pronounced in substrates that have much kaolinite or are of coarse texture (Fig. 2b). Results with these substrates show a more or less rapid loss of elasticity.

Increases of G'' in Phase 2 of all untreated Typic Hapludox samples, either saturated or predrained, indicate frictional heat. It results from the reorientation of particles, e.g., kaolinite platelets, packages, or single grains. Consequently, deformation {gamma} increases, because frictional heat needs to be generated to cause a structural breakdown in Phase 3. In general, viscoelastic and cohesive substrates also show this micromechanical behavior. They typically react with a temporal delay (Mezger, 2002).

Intersections of G' and G'' of Typic Hapludox samples (Fig. 2b–2e) occur within a range of 50 to 80% deformation under saturated conditions and 70 to >90% under predrained conditions (Fig. 3a–3d). In Fig. 3b and 3d, intersections of G' and G'' of untreated clay-rich Hapludox samples are absent, and G' is higher than G'', which indicates a very rigid character of the substrates, despite a relatively rapid decline of G' and G'' in the beginning.

Leaching effects were observed in almost every case, depending on differences in texture, organic matter, clay, or Fe oxide contents at different water contents. With regard to the sandy Typic Hapludox soil (Fig. 2b), SOM-leached samples show few deviations compared with untreated samples. The effect of frictional heat decreased gradually from untreated > SOM-leached > Fed–leached substrates. This effect also can be derived from the curve behavior, which changes from a flat s-character (untreated) to a straightened graph (leaching), followed by a decreasing elastic behavior, a lowering of G' and G'' levels, and a reduced distance between graphs of G' and G'' in the plateau phase.

Characterization of the Mineral Composition by Scanning Electron Microscopy–Energy Dispersive Scan Analyses and X-ray Diffractometry
To explain the rheological findings, SEM–EDS analyses were performed. Based on the visual findings of the SEM micrographs, a typical smectitic (HAC) structure is given for the Typic Calciudert soil (Santana do Livramento; Fig. 4a). X-ray diffractometry graphs confirm these findings. High-feldspar compounds, as well as kaolinite, occur with smaller amounts of Fe (hydr)oxides.


Figure 4
View larger version (146K):
[in this window]
[in a new window]

  		Fig. 4. Scanning electron micrographs of (a) Santana do Livramento Typic Calciudert, Rio Grande do Sul (RS), untreated; (b)–(d) Cruz Alta, RS, sandy Typic Hapludox: (b) untreated, (c) soil organic matter (SOM) leached, and (d) dithionite-soluble iron (Fed) leached; (e)–(g) Cruz Alta, RS, clayey Typic Hapludox: (e) untreated, (f) SOM leached, and (g) Fed leached; (h)–(j) Santo Ângelo clayey Typic Hapludox, natural forest: (h) untreated, (i) SOM leached, and (j) Fed leached; and (k)–(n) Santo Ângelo, clayey Typic Hapludox, under no-tillage: (k) untreated, (l) SOM leached, (m) Fed leached, and (n) detail of (m).

 
Untreated kaolinitic samples of the Typic Hapludox soils from Cruz Alta and Santo Ângelo show the phenomenon of pseudosand aggregation (Fig. 4b, 4e, 4h, and 4k). Surfaces are covered with kaolinite and Fe (hydr)oxides, whereas samples from Cruz Alta are dominated by hematite. This is in contrast to a greater number of goethite compounds in Hapludox samples from Santo Ângelo. Additionally, single basalt grains (Fig. 4g, after the sodium dithionite treatment) are stuck together by fine roots (Fig. 4h and 4k), which merge into microcracks (Fig. 4c) and cavities on the surface and kaolinite piles (Fig. 4d). In Fig. 4j, a single crystalline structure of magnetite is shown. Both samples from Santo Ângelo, which were treated with sodium dithionite, feature this structure. Light microscopy under polarized conditions gave the same results, because magnetite has an opaque character. Furthermore, EDS analyses corroborated these findings. High contents of elementary Fe were detected with EDS, which supports the existence of magnetite as well as various amounts of Al and Si in the surrounding structures, i.e., kaolinite. In addition, halloysite, which is characterized by high Al contents (confirmed by EDS analyses), is shown in detail in Fig. 4n for the clayey Typic Hapludox soil from Santo Ângelo.


   DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Rheological data confirm that organic matter content and Fe oxides influence aggregate stability and shear behavior under oscillatory loads. The SEM micrographs of differently treated samples (e.g., samples that have the organic matter removed) show transitions from complex pseudosand aggregates to even surfaces of single, disconnected particles. Binding mechanisms that strengthen aggregates decrease stepwise from untreated, to SOM-leached, to Fed–leached samples. These results are in agreement with the findings of Lamotte et al. (1997), Amado et al. (2001), Imhoff et al. (2002), and West et al. (2004), who have investigated friability of such soils. Additionally, Muggler et al. (1999) discovered diverging effects of Fe oxides and organic matter in Oxisols. They demonstrated that aggregation by Fe oxides is evident in soils that developed on Tertiary sediments compared with those that have rock-saprolite as parent material. Our findings support the idea that the effect of Fe (hydr)oxides seems to be indirect, through binding with organic matter. Aggregate formation and strengthening are strongly enhanced, however, by the remobilization of Fe and the conversion of ferrihydrite to hematite (Ohtsubo et al., 1991). Remnants of magnetite in sodium-dithionite-treated Typic Hapludox soils from Santo Ângelo (natural forest and no-tillage) show another aspect of pseudosand texture. An assemblage of single grains (e.g., formed by silt) defines this phenomenon, as well as bridges of Fe oxides (magnetite) or packages of kaolinite, which may function as single particles with respect to shear behavior. Altered curve progressions and phase characteristics derive from this instance, especially in Phases 1 (plateau phase) and 2 (intersection of G' and G''). In general, such assemblages (Fig. 4b, 4c, 4e, 4h, and 4k) lead to a more elastic behavior than single particles. Differences in shear behavior of such aggregates, in comparison to single grains, can, therefore, be parameterized and defined.

Syers et al. (1971) investigated phosphate sorption parameters of representative soils from Rio Grande do Sul and also reported the mineralogical characteristics of soils from Cruz Alta and Santo Ângelo, among other sites. They pointed out that SEM–EDS observations showed the appearance of interconnected voids between skeleton basalt and sand and silt grains coated with clay (kaolinite packages, smectitic layer structures) or Fe (hydr)oxides. In this context, the phenomenon of wall-shaped bridges, shown by Lamotte et al. (1997), should be mentioned. They produced SEM micrographs of hard-setting, Fe-(hydr)oxide-rich soils from North Cameroon, which showed interconnected voids of broken surfaces and small fragments (compare Fig. 4c and 4f). The effects of H2O2- and dithionite-treated samples, namely the gradual reduction of functioning connectors or bridges between single grains, can be visualized and compared directly with untreated samples. These results are also consistent with the findings of Tisdall and Oades (1982), whonoted, on the one hand, defined microaggregates in general as <250 µm and, on the other hand, the strengthening effect of organic matter for soil aggregation. Oades and Waters (1991) not only introduced a hierarchical system for aggregates or associations of clay aggregates, but also described a typical assemblage of kaolinite crystals. Thus, based on these SEM findings, results of the amplitude sweep tests confirm the assumption of stabilization effects caused by Fe (hydr)oxides and SOM, when one considers the shear behavior and viscoelastic characteristics. Confirmation also is reflected in derived values of {gamma}L and {tau}y (Table 2).


   CONCLUSIONS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
Amplitude sweep tests, made with a parallel-plate rheometer, revealed microstructural changes in Oxisols, which are influenced by Fe (hydr)oxide formations, textural differences (clay mineralogy), or water content. In agreement with Hypothesis 1, results show that Fe (hydr)oxides and SOM have an influence on microstructural stability.

Curve characteristics, as well as single parameters like G', G'', the calculated LVE deformation range, and {gamma}L, enabled the quantification of dissipating elasticity (loss of rigidity) in soils on a microscale (particle–particle scale). By conducting amplitude sweep tests, significant differences between these characteristics in Hapludox and Calciudert soils were found. Untreated Typic Calciudert samples had relatively low values of {gamma}L and {tau}y, as shown by these tests. This is because differences in clay mineralogical properties have a major influence. Microstructural stability of smectitic substrates was less than that of kaolinitic ones as presented in the Hapludox soils, but they maintained a more or less stable level of viscoelasticity. The dissipation of elasticity occurred gradually. Sliding (platy or aligned particles) behavior is contrasted to a turbulent (rotund shape) shear behavior. Micromechanical characteristics under oscillatory conditions were parameterized by rheological data and complemented and visualized by SEM–EDS findings.

Visual findings from SEM micrographs indicated transformations in aggregation due to different treatments: pseudosand aggregates in untreated samples, looser structures in SOM-leached samples, and single-particle conditions in dithionite-treated Typic Hapludox samples. Micropores and microcracks increased in size until unconnected conditions were achieved.

Consequently, problems regarding upscaling—from tactiles, grains, (micro)aggregates, or, as shown in this study, pseudosand structures, up to a meso (aggregate) or even macro scale (undisturbed samples)—can be resolved. A trend of (dis)aggregation sequence can be distinguished, which is either influenced by given physicochemical properties, especially different clay contents and mineralogical compounds, or by laboratory-induced effects due to H2O2 or sodium dithionite treatments. The interaction of both SEM–EDS and rheological findings provides information about micromechanical processes and leads to a better understanding of shear behavior under oscillatory loads.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.

Received for publication June 15, 2006.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 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 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 Google Scholar
Google Scholar
Right arrow Articles by Markgraf, W.
Right arrow Articles by Horn, R.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Markgraf, W.
Right arrow Articles by Horn, R.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Markgraf, W.
Right arrow Articles by Horn, R.
Related Collections
Right arrow Soil Physics

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
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome