| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Faculty of Natural and Agricultural Sciences The University of Western Australia 35 Stirling Highway Crawley Western Australia 6009
jquirk{at}agric.uwa.edu.au
The paper by McBride and Baveye (2002) lacks balance in its presentation and is not really relevant to soil science. We are informed that there is both support and opposition in the literature for the Coulombic attractive theory of Sogami and Ise (1984), which they used to explain the formation of a liquid crystalline phase in dispersed clay suspensions (McBride and Baveye, 2002, p.1216). However the authors, surprisingly, do not refer to two of Overbeek's papers ( Overbeek, 1987; 1993) in which he exposes errors in the theory. More recently Petris and Chan (2002) have successfully treated the formation of a liquid crystalline phase without recourse to the Coulombic attractive theory. In the conclusions we are told that the phenomenon ‘may not occur, or may be difficult to detect in real colloids.’ Thus the paper concerns a theory that the authors tell us is questionable and which may not apply to real colloids, that is to clay-particle interaction in a soil where the total water content, even for clay-textured soils, would not exceed 50% by weight at 10 kPa suction (Holmes, 1955); so the organization of clay crystals and the separation of their surfaces in a soil contrasts markedly with that in dilute clay suspensions discussed by the authors.
The authors, although referring to Quirk (1994), apparently did not appreciate that this review rendered their considerations of little relevance to ‘real’ soils in which the clay crystals exist in compound particles, composed of many plate-shaped crystals in parallel or near parallel alignment, which Aylmore and Quirk (1960) described as "clay domains"; expressed succinctly soils are condensed colloid systems. When Ca is the dominant exchangeable ion the clay domains are stable entities in which the relationship between the clay crystals within a domain can be represented by a three-crystal model in which one crystal separates two other crystals in such a manner as to form a slit-shaped pore and smaller pores where the crystals overlap (see Fig. 5 in Quirk, 2001). The slit-shaped pores have surface separations that are determined principally by the distribution of crystal thicknesses; Aylmore and Quirk (1967) have shown that there is a distribution of pore sizes for Willalooka illite (1.6 x 10-5 m2kg-1) with a median size of 4.5 nm. Water is able to enter the overlap pores for which the size would be of the magnitude of 1 nm.
Within the slit shaped pores repulsive diffuse double layer (ddl) pressures would act to cause the crystals to separate and as a result the domains as a whole would swell. In the overlap pores, two types of attractive pressures act to restrain this swelling and are the basis for the stability of clay domains. These pressures are the classical van der Waals pressure between macroscopic objects with pairwise interaction and ion-ion correlation pressures. This latter pressure can be understood by considering the limited crystalline spacing of Ca-montmorillonite, which even in distilled water is limited to a d(001) value of 1.9 nm (Norrish and Quirk, 1954). The lamellae surfaces are separated by 1 nm and it can be shown that the concentration of the Ca ions in this space approaches 3 M, creating strong electrostatic interactions (correlation) between the divalent ions that generate a strong attractive pressure; Kjellander et al. (1988) calculated that the attractive pressure resulting from van der Waals forces and ion-ion correlation between Ca-montmorillonite lamellae as 1.5 MPa for a separation of 1 nm (see Fig. 7 in Quirk, 1994). Because this attractive pressure exceeds the repulsive pressure caused by ion hydration (Slade and Quirk, 1991), the Ca-montmorillonite lamellae reside in a potential minimum or well. The attractive pressure minus the hydration repulsion pressure is described as the resultant attractive pressure. For Na-montmorillionite the hydration pressure dominates and causes extensive crystalline swelling in solutions more dilute than 0.3 M NaCl (Norrish and Quirk, 1954).
Returning to the three-crystal model, the interplay between the repulsive ddl pressure in the slit-shaped pores and the resultant attractive pressure in the overlap pores is of special interest when the exchangeable Na percentage (ESP) increases because the attractive pressure, resulting from electrostatic interactions, in the region of crystal overlap becomes less as Na+ replaces Ca2+ and in the slit-shaped pores the repulsive pressure increases; so that as the concentration of the external solution is decreased for a given ESP value the total repulsive force (pressure multiplied by area of slit-shaped pore) eventually exceeds the total attractive force (resultant pressure multiplied by the area of the overlap pores) and as a result the clay domains start to swell. It follows that this swelling can be prevented if the external electrolyte concentration is sufficiently large to reduce the swelling pressure associated with slit-shaped pores in accordance with the ddl theory. This is the basis for the threshold concentration concept (Quirk and Schofield, 1955; Quirk, 2001).
Quirk and Schofield (1955) studied the permeability of a loam-textured soil in relation to ESP and the electrolyte concentration of the percolating solution. Using these results Quirk (see Quirk, 1994) obtained the following relationships
 | [1] |
 | [2] |
The threshold concentration corresponds to the beginning of the swelling of clay domains and the turbidity concentration corresponds to the start of the disruption of clay domains. This disruption, as judged by the degree of turbidity, increases with increasing ESP values and results from the dominance of the total repulsive force over the total resultant attractive force and is assisted by the thermal motion of the particles. This spontaneous dispersion takes place at much smaller concentration than that required to effect the flocculation to dispersion transition in a suspension because in preparing a suspension the clay domains are disrupted whereas within a soil the concentration of the percolating solution has to be sufficiently small for the repulsive pressure to remove the clay crystals from within the potential minima in which they reside in a clay domain. In a suspension edge to face interactions are an important contribution to the flocculation process as indicated by the effect of dispersing agents or organic moieties, adsorbed on the edge face of crystals in increasing the concentration required for flocculation. This can be represented thus:
 |
While the success of the three-crystal model in providing a general description of the behavior of sodic soils and the nature of soil microporosity is satisfactory, more detailed consideration of the character of overlap pores is justified in view of their crucial role. It would not be expected that the degree of conformity of contiguous illite crystals would approach that of two montmorillonite lamellae as the degree of alignment and other factors would be involved such as the steps on crystal surfaces. Also the relative contributions of ion hydration and ion-ion correlation to short-range interactions is yet to be resolved (Kjellander et al., 2001).
Another feature of clay domains is that they are in random array in a soil and a priori it would be expected that discontinuities would follow the surface of the domains. Murray and Quirk (1990) found that the surface area of clays and soils obtained from low temperature N2 adsorption (using the BET equation) and desorption (using the Kelvin equation) isotherms, show remarkable agreement indicating the ready ingress and egress of the N2 molecules to and from the clay surfaces via what are appropriately called intrinsic discontinuities.
Intrinsic discontinuities are significant with respect to clay and soil water relations. When Ca ions balance the surface charge, the limited swelling of the clay domains is complete at about 5 MPa. At suctions smaller than this value the intrinsic discontinuities progressively expand as the suction is progressively reduced when energy stored in the drying cycle is progressively released. This mechanism is the principal reason for clay and soil swelling and being a mechanical process is independent of electrolyte concentration as demonstrated by Ca-Willalooka illite, which attains a water content of 0.46 kg kg-1 over the range of CaCl2 concentrations form 1 M to distilled water (Aylmore and Quirk, 1962). In light-textured soils the clay swelling is accommodated in the coarse porous structure. Thus the basis for soil water content-suction relationships is related to the presence of clay domains.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
