Research Triangle Institute
Research Triangle Park, North Carolina
Park Ridge, New Jersey, U.S.A.
Gothard House Publications
Gothard House, Henley-on-Thames
Clay liners are composed of layers of cohesive soil, engineered and compacted to form a barrier to liquid-migration. From an engineering standpoint, soil has been defined as unconsolidated accumulations of solid particles produced by the physical and chemical disintegration of rocks (ASTM, 1985) or all materials in the surface layer of the Earth’s crust that are loose enough to be moved by a spade or shovel (Winterkorn and Fang, 1975).
Soil may be viewed as a three-phase system composed of solids, liquids, and gases. The solid phase is composed of inorganic and organic particles of varying shapes and sizes. The liquid phase is usually an aqueous electrolyte solution. The gaseous phase is basically air with variations in composition resulting from biological activity and chemical processes in the soil. Soils are normally characterized by the size and composition of their solid (particulate) components, with air- and water-filled voids considered together as porosity. However, the relative amounts of air and water (usually expressed as the degree of saturation) also influence soil behavior.
The term clay can be defined in several ways. Clay can refer to all soil particles less than a given size, usually 2 μm (Mitchell, 1976). In soils, this particle size range is composed of clay minerals and other components. Clay minerals give a clay soil its plastic and cohesive properties. Other components of the clay size fraction include nonclay minerals (e.g., marl and chalk), amorphous material, and organic material.
Geotechnical engineers use the term clay to refer to soils that contain enough clay size particles to affect their behavior (Holtz and Kovacs, 1981). Because of its emphasis on the engineering properties of clay as a liner material, this document uses the term clay to refer to clay soil, with clay mineral being used when referring to mineralogy and clay size being used when referring to particle size.
This chapter discusses the characteristics of clay soils and the clay minerals that are important soil constituents. It is intended to give the reader a brief overview of these materials, to present some basic definitions, and to discuss the properties of clay soils and clay minerals that influence the performance of clay liners. The formation and occurrence of low-permeability, clay-rich soils are also discussed. For a more thorough discussion of clay mineralogy or of the geotechnical behavior of soil, the reader is referred to Grim (1962, 1968), Mitchell (1976), and Perloff and Baron (1976).
Clay minerals are hydrous silicates, largely of aluminium, magnesium, and iron, that, on heating, lose adsorbed and constitutional water and yield refractory material at high temperatures. Plasticity is characteristic of clay minerals and is largely due to an affinity of the clay surface for water, resulting from a net negative charge on the surface of a clay particle that causes it to adsorb water and other polar fluids. This net negative surface charge results from defects in the clay mineral crystal structure an from surface chemical reactions, as described below. Because of their electrochemical surface activity and high surface area (resulting from their small size and lamellar shape), clay minerals can profoundly affect a soil’s engineering behavior, even when present in small quantities. As the clay content of a soil increases, the influence of the clay fraction on its behavior also increases. The strong influence of clay minerals on soil behavior can be illustrated by the addition of bentonite to a granular soil. Bentonite is a clay material composed largely of the clay mineral sodium montmorillonite. An addition of only 2 to 3 weight percent bentonite can reduce a soil’s permeability when it is compacted by 2 to 3 orders of magnitude (Kozicki and Heenan, 1983). The influence of clay minerals on a soil’s behavior increases with increasing clay content to a range of 33 to 50 percent, at which point the nonclay size material is essentially floating in a clay matrix and has little effect on the engineering behavior of the soil (Mitchell, 1976; Holtz and Kovacs, 1981).
Most clay minerals have a sheet-like layered crystalline structure and thus fall into the phyllosilicate mineral family. Exceptions are the clay minerals attapulgite, palygorsite, and sepiolite, which have structures composed of double chains of silica tetrahedra. These minerals are not common in soils and are not discussed further in this document.
Clay mineral sheet structures consist of two different layer types, one composed of tetrahedral units and the other of octahedral units, that are arranged in different sequences to form the different clay minerals. The tetrahedral unit is composed of silica tetrahedra in which four oxygens surround a silicon atom in tetrahedral coordination (Figure 2-1).
The octahedral sheet, which is made up of cations octahedrally coordinated with oxygen (Figure 2-2), occurs in two forms. If the cation is trivalent, only two-thirds of the possible spaces in a layer are filled and the structure is dioctahedral. The most commonly occurring dioctahedral sheet in clay minerals is the gibbsite sheet, in which the cations are aluminum. If the cation in the octahedral sheet is divalent, all of the available cationic spaces are filled and the structure is termed trioctahedral. The most commonly occurring trioctahedral sheet in clay minerals is the brucite sheet, in which the cations are magnesium.
Isomorphous substitution, or the substitution of different, similar-sized cations for those present in the ideal crystal structure without a change in structure, is common in clay minerals and is an important factor in their behavior. Common cation replacement in clay minerals includes aluminium (Al+3) for silicon (Si+4), magnesium (Mg+2) for aluminium (Al+3), and ferrous iron (Fe+2) for magnesium (Mg+2 ) in the ideal tetrahedral and octahedral sheets described above. Isomorphous substitution in clay minerals results in a charge deficiency in the crystal structure and a net negative charge on the mineral’s surface.
The variety of cation substitutions, both in and between the crystalline sheets, and the intergrowth of layers of different character results in the diversity of actual clay minerals. However, despite the large number of clay minerals, only a few are important soil-forming constituents. Table 2-1, (from Mitchell, 1976) summarizes important chemical and physical characteristics of the clay minerals that commonly occur in soils.
This section describes the basic structural makeup of common clay mineral groups. Although each clay mineral has a definite "ideal" structure, many naturally occurring clays are complex and do not fit the ideal formulas described herein. Mixed-layer clays can occur, with crystals containing structural units of more than one clay mineral group. In addition, soils composed of a single clay mineral or clay mineral group are relatively rare; multimineral soils are more commonly encountered in most areas. In a study of 137 soils across the United States, more than one clay mineral occurred in about 70 percent of them (Lambe and Martin, 1953-1957). Therefore, it is not possible to predict soil behavior accurately by assuming that only one clay mineral predominates through soil material. For more information on clay minerals and their complex natures, the reader is referred to Grim (1968) and Van Olphen (1963).
All clay minerals, except those with chain structure, may be roughly categorized into four groups based on the height of the unit cell, the composition of the sheets, and the kind of intersheet bonding that forms the layers of unit cells. These groups are kaolinite, smectite, illite (or mica-like), and chlorite. This grouping ls convenient since members of the same group have comparable engineering behavior (Mitchell, 1976). The following subsections describe these groups and the mineral characteristics that are important determinants of the engineering behavior of clay.
The basic structural unit (unit cell) of the kaolinite group is a 1:1 arrangement of a silica tetrahedral sheet and an alumina (gibbsite) octahedral sheet (Figure 2-3).
In the tetrahedral sheet, the tips of the silica tetrahedra all point toward the center of the unit. The oxygen atoms at the tips of the tetrahedra are common with one of the planes of oxygens in the octahedral sheet and compose two-thirds of the octahedral oxygens. The remaining positions in this plane are occupied by hydroxyls that are located directly below each hexagonal hole in the network formed by the bases of the silica tetrahedral (Figure 2-3). The kaolinite basal spacing is 7.2 Angstrom.
Minerals in the kaolinite group are composed of stacks of the 1:1 structural units (unit cells) described above. These unit cells are held together by van der Waals forces and hydrogen bonding between the tetrahedral sheets and the octahedral sheets of adjacent unit cells. These bonds are strong enough to preclude the introduction of water between the unit cells and thus any interlayer swelling.
184.108.40.206.1 Kaolinite – Kaolinite is the most common mineral in this group and consists of stacks of 1:1 unit cells comprised of silica tetrahedral and gibbsite (Al) octahedral sheets. The stacks generally range from 0.05 to 2 μm in thickness and can attain thicknesses up to 4,000 μm; the stacks can range from 0.1 to 4 μm laterally. The specific surface area of kaolinite is on the order of 10 to 20 m2/g of dry clay.
A small net negative charge on kaolinite particles results in a cation exchange capacity of 3 to 15 meq/100 g. This charge has been attributed to a small amount of isomorphous substitution in the silica or gibbsite sheets, replacement of exposed hydroxyl hydrogens by exchangeable cations, broken bonds around particle edges, or diffuse charges resulting from the large size and surface accessibility of 0-2 and OH- molecules (Mitchell, 1976; Winterkorn and Fang, 1975). The stacked crystal structure of kaolinite results in a blocky form for this clay mineral and a larger size and lower surface-to-volume ratio than other clay minerals. This low surface area, combined with the relatively small negative surface charge, results in kaolinite being the least electrochemically active and least plastic clay mineral.
Because of the blocky structure of kaolinite particles, crystal edges of this mineral group comprise 10 to 20 percent of the total crystal area (Theng, 1974) and therefore exert a stronger influence on the electrochemical behavior of these minerals than do the crystal edges of smectites or illites (crystal edges comprise 2 to 3 percent of total crystal area for montmorillonite; Theng, 1974). Broken bonds on these edges result in unsatisfied valences that can be satisfied by cation or anion adsorption. However, unlike the structurally generated negative charges on the platy crystal surfaces, these charges are affected by the pH of the environment. Evidence suggests that the edges are positively charged at low pH and negatively charged at high pH. This results in kaolinite having a low cation exchange capacity at low pH and higher cation exchange capacity at high pH (see Section 2.3.1). Kaolinite has a higher anion exchange capacity than most clay minerals. This may result from the presence of replaceable hydroxyl ions on the outside of structural sheets. Kaolinite thus has the ability to fix certain negative ions (Deer et al., 1966).
Compared with other clay minerals, kaolinite has a lower affinity for water, has a lower dispersivity (see Section 2.3), and does not achieve as low a permeability upon compaction. On the other hand, because it is not as electrochemically active, its behavior may be less affected by chemicals than other clay minerals. Thus, a kaolinitic clay liner may have a higher permeability than liners composed of other clays, but the permeability of a kaolinitic clay liner may not be as sensitive to changes in moisture content or to chemical attack.
220.127.116.11.2 Halloysite – Halloysite is another kaolinite group mineral that is a common soil constituent in some areas. This mineral occurs in two forms: a nonhydrated type with a structural composition similar to kaolinite and a hydrated form with a single layer of water interposed between unit kaolinite layers (Figure 2-3). This layer increases the basal spacing to 10.1 Angstrom, compared with 7.2 Angstrom for nonhydrated halloysite and kaolinite. Partially hydrated halloysite (metahalloysite) with basal spacing from 7.4 to 7.9 Angstrom can also occur. The interlayer water molecules in hydrated halloysite are believed to be in a rather flat hexagonal network linked to each other and to adjacent halloysite layers by hydrogen bonding.
The hydrated form of halloysite occurs in cylindrical tubes of overlapping kaolinite sheets. The outside diameters of the tubes range from 0.05 to 0.20 μm, with a median value of 0.07 μm, and range in length from a fraction to several micrometers. The specific surface area of halloysite ranges from 35 to 70 m2/g (Mitchell, 1976).
Because of the interlayer water sheet in hydrated halloysite, intercalation (introduction between the unit cells) of chemicals can occur. This also results in a slightly higher cation exchange capacity for hydrated halloysite (5 to 40 meq/100 g) than for kaolinite (3 to 15 meq/100 g). Halloysite also may be more affected by chemicals than kaolinite.
The interlayer water in halloysite is easily removed during drying, and this dehydration is irreversible. Because of this phenomenon, soil engineering tests on air-dried samples can give different results than those performed on samples at the original field moisture content. For this reason, it is especially important that laboratory tests on soils with appreciable halloysite content be carried out on samples at the original field moisture content (Holtz and Kovacs, 1981; Hilf, 1975).
18.104.22.168.1 Illite – Illite is an important constituent of clay soils and has been described by Mitchell (1976) as "perhaps the most commonly occurring clay mineral found in soils encountered in engineering practice." Illite has almost the same crystalline structure as muscovite mica. This structure is comprised of a silica-gibbsite-silica sandwich, with the tips of the silica tetrahedra pointing toward the octahedral gibbsite sheet and the oxygens at the tips being common with the octahedral sheet (Figure 2-4).
Isomorphous substitution of aluminium for silicon in the tetrahedral sheet results in a negative charge at the surface of these layers. This charge is balanced by potassium, cesium, and ammonium ions between the 2:1 layers; these ions fit tightly in the 1.32-Angstrom-radius holes in the bases of the silica sheet and as a result are fixed in position and are not exchangeable. Illite differs from muscovite in having less isomorphous substitution in the tetrahedral sheet, a lower negative surface charge, and a lower amount of potassium between the layers. The stacking of illite layers is also more random, and illite occurs with a much smaller particle size than muscovite.
In terms of properties important to clay liner performance, illite lies between kaolinite and the smectite clay minerals. Although extensive isomorphous substitution results in a net negative charge on the clay mineral surface, the fixed potassium cations balance the charges and strongly bond adjacent 2:1 sheets together. As a result, illite has intermediate values for surface area (65 to 100 m2/g), cation exchange capacity (10 to 40 meq/100 g), swelling index, and activity. It is also intermediate in its reaction to chemicals. Because of the strength of the interlayer potassium bonding, the basal spacing of illite remains at 10 Angstrom when it is exposed to polar liquids (Mitchell, 1976). The potassium ions effectively prevent the intercalation of water, organic liquids, and other cations (Deer et al., 1966).
22.214.171.124.2 Vermiculite – Vermiculite is a fairly common mineral in clay soils and usually occurs with other clay minerals. Vermiculite has a 2:1 structure with a poorly organized octahedral sheet sandwiched between two silica tetrahedral sheets (Figure 2-4). The octahedral sheet contains iron and magnesium ions.
As with illite, isomorphous substitution of aluminum for silicon is extensive in the tetrahedral sheet, resulting in a net negative charge on the crystal surface. This positive charge deficiency ls larger than that of the smectite minerals (see Section 126.96.36.199) and is usually balanced by interlayer layers of divalent cations and water. This larger charge deficiency results in vermiculite having the highest cation exchange capacity of all clay minerals (Deer et al., 1966). The most common interlayer cations in vermiculite are magnesium and, to a lesser extent, calcium.
The amount of water that is intercalated in vermiculite is less variable than that in smectite and usually is limited to two layers of water molecules. The interlayer spacing is therefore fairly constant for vermiculite but varies to some extent depending on the cations present between the layers. Vermiculites can absorb organic liquids between their layers but take up less than the smectite minerals (Deer et al., 1966). The primary specific surface area for vermiculite ranges from 65 to 100 m2/g. This is within the range reported for montmorillonite and, as with montmorillonite, the secondary (interlayer) surface area can reach very high values (870 m2/g) (Mitchell, 1976).
Chlorite minerals in clay soils are almost always found in association with other clay minerals. Chlorites are composed of 2:1 layers of silica (tetrahedral sheets surrounding a gibbsite or brucite octahedral sheet, with another octahedral sheet between the 2:1 mica layers (Figure 2-5). Chlorites thus may be termed 2:1:1 clay minerals.
Chlorites can have isomorphous substitution and may be missing a few of the octahedral sheets between the 2:1 layers. This can result in some swelling from water uptake between the layers. Chlorites are less active than the smectites, have a cation exchange capacity similar to illite (10 to 40 meq/10 g), and may be similar to illite in engineering behavior.
The smectite group of clay minerals includes 2:1 minerals whose unit cell is composed of an octahedral sheet sandwiched between two silica tetrahedral sheets. The bonding between these 2:1 layers is by van der Waals forces and cations that may be present to balance out structural charge deficiencies in the 2:1 layer. This bonding is weak, and, as a result, the layers are easily separated by adsorption of water or other polar liquids. Thus, the interlayer spacing of smectites can vary from 9.6 Angstrom to complete separation, and this results in the high swelling behaviour and high activity of these clay minerals.
The smectite minerals may be divided into two groups, based on the composition of the octahedral sheet. The montmorillonites have a dioctahedral, aluminium-based (gibbsite) octahedral sheet; the saponites have a trioctahedral magnesium-based (brucite) sheet. Only montmorillonite is commonly found in soils. The saponites are relatively unimportant as soil constituents and are not discussed further in this document.
188.8.131.52.1 Montmorillonite – Montmorillonite is a 2:1 clay mineral with a dioctahedral gibbsite sheet sandwiched between two silica tetrahedral sheets (Figure 2-5). Extensive substitution of magnesium and other cations for aluminium and aluminium for silicon results in a charge deficiency of 0.5 to 1.2 (usually 0.66) on the unit cell (Mitchell, 1976). Most of the substitution in montmorillonite occurs in the octahedral gibbsite sheet, usually one magnesium for every sixth aluminium. This results in a charge on the mineral surface that is more diffuse or evenly spread than that of vermiculite, which has mostly substitution of aluminium for silicon in the outer, tetrahedral layers (Deer et al., 1966; Winterkorn and Fang, 1975). The charge deficiencies on the montmorillonite unit cells are balanced by exchangeable cations between the unit cells, and, as a result, montmorillonite exhibits high cation exchange capacity (generally 80 to 150 meq/1O0 g).
The bonding forces between unit cells of montmorillonite are weak, and water and polar fluids can easily penetrate between the unit cell layers. As a result, montmorillonite particles are very small and can be dispersed to sheets of unit cell thickness (10 Angstrom) in water (Mitchell, 1976). The specific surface area of montmorillonite is very high, with a primary surface area of 50 to 120 m2/g and a secondary surface area (including interlayer surfaces) of 700 to 840 m2/g. Because of its high specific surface and tendency to adsorb interlayer water, montmorillonite is very susceptible to swelling and is the most active of the clay minerals. Montmorillonite is especially sensitive to alteration by chemical attack.
The type of cations occupying the interlayer spaces strongly influences the behavior of montmorillonite. The most commonly occurring inter-layer cation is calcium, a divalent cation. Like vermiculite, calcium-montmorillonites usually take up two layers of water between the unit cell layers (Deer et al., 1966). This results in limited swelling to a maximum interlayer spacing of 19 Angstrom (Theng, 1974). However, when sodium is the interlayer cation as occurs in the Wyoming bentonites (see 184.108.40.206.2) the amount of interlayer water is not so limited and the interlayer spacing can range from 10 Angstrom (oven-dry) to over 50 Angstrom (Theng, 1974) This results in high swelling, which is characteristic of sodium-montmorillonite; it can expand to 13.8 times its dry volume when fully hydrated.
220.127.116.11.2 Bentonite – Bentonite ls not a clay mineral. It ls a rock (or clay deposit; composed largely of the clay mineral montmorillonite. The swelling and dispersive properties of this mineral give bentonite the ability to lower the permeability of a soil, even when added in small quantities (e.g., 1 to 3 percent by weight). The swelling capacity of bentonite depends on its sodium-montmorillonite content. Low-swelling bentonite has significant quantities of calcium-montmorillonite, which, because of limited interlayer water uptake, does not swell to the extent of sodium-montmoriltonite. High-swelling sodium bentonite has a liquid limit of 500 percent or more and can swell 15 to 20 times in volume. Calcium bentonite will increase in volume O to 5 times when wetted with water; this swelling capacity has been reported to increase 700 to 1,000 percent by treating calcium bentonites with a 0.25-percent solution of Na2CO3 (Fisher, 1965).
Bentonite is formed by the weathering of volcanic ash. Environmental conditions favorable to sodium bentonite formation are semiarid climate and alkaline soil and groundwater. The type locality for bentonite is Wyoming, and most sodium bentonite comes from the western United States and Canada (Hosterman, 1985). Calcium or low-swelling bentonite is mainly obtained from deposits in the Gulf Coastal Plain formed from weathering of volcanic ash deposits (Hosterman, 1984).
This section presents some of the factors that influence the formation and occurrence of clay soils in the United States. It is intended to help the reader comprehend the complexity of soil-forming processes and the degree of heterogeneity and variability that may be expected in naturally occurring soils that may be used for clay liners. It is not intended to be a complete treatise on the subject.
Clay minerals are products of weathering or hydrothermal alteration, with different minerals resulting from differences in physical-chemical conditions and differences in parent material (Deer et al., 1966). Clay minerals may be found in their place of origin or may be transported and deposited in sediments. In general, acid conditions favor the removal of cations from the soil and kaolinite formation, and alkaline conditions favor the formation of other clay minerals, with the predominant type(s) of cation influencing the clay mineral species. A brief discussion of clay mineral formation follows. A more complete discussion may be found in Grim (1968), Keller (1964), and Weaver and Pollard (1973).
Kaolinite mineral deposits are formed primarily by the weathering of alkali feldspars and other silicate minerals common to sialic rock types such as granites and quartz diorite. Kaolinite occurs in residual, hydrothermal,