Since structural Fe
2+
increases the interaction between H
2
O and the clay mineral
surface, these thermodynamic arguments predict that swelling in reduced smectite
should be greater than in oxidised smectite. If this is true, then why are the observed
effects just the opposite (Foster, 1953, 1955; Egashira and Ohtsubo, 1983; Stucki et al.,
1984b, 2000; Lear and Stucki, 1985; Gates et al., 1993). The answer is found by
comparing the studies of Viani et al. (1983) and Wu et al. (1989), who discovered that
two types of interlayers are possible in swelling smectites: fully expanded and fully or
partially collapsed. At any given swelling pressure, the distance between the fully
expanding layers is the same, regardless of the water content of the clay mineral at that
pressure. So the differences in water content for two clay minerals at the same swelling
pressure occurs because of a difference in the fraction of layers that are fully expanded
relative to the fraction that are partially or fully collapsed, rather than from the layers
expanding to different distances. They described these relationships by the expression
ln P þ 1ðÞ¼ln b þ
a
l
ð5Þ
where a and b are constants and l is the interlayer distance. Notice that the swelling
pressure, P, is a single-valued function of interlayer distance.
Reduction of octahedral Fe
3+
causes more of the clay mineral layers to collapse
compared to the oxidised state, thereby removing those layers from the pool of fully
expanded layers (Wu et al., 1989). The overall capacity of the clay mineral to adsorb
water on a mass basis is thus diminished. This observation compares well with other
studies showing an increase in cation fixation as the reduced state of the clay mineral
increases. Chen et al. (1987) observed an increase in K fixation in agricultural soils as
the amount of structural Fe
2+
in the constituent clay minerals increased, and Khaled
and Stucki (1991), Lear and Stucki (1987), and Shen and Stucki (1994) confirmed
this principle in standard reference clay minerals, indicating that layers are indeed
collapsing around these cations. The extent of cation fixation by the reduced smectite
depends inversely on the hydration energy of the cation (Khaled and Stucki, 1991).
Another consequence of the increased interaction between reduced smectite sur-
faces and interlayer H
2
O is that the hydration energy of the surfaces exposed to H
2
O
should increase, even though the net water holding capacity can decrease for the
reasons just explained. Stucki et al. (2000), in a study of the effects of organic cations
on clay mineral swelling, demonstrated that this is indeed the case. The quaternary
ammonium cation trimethylphenyl ammonium (TMPA
+
) was exchanged onto the
oxidised and reduced smectites, then the water retention curve was obtained. Water
retention curves were also obtained for the Na
+
-exchanged analogues (Fig. 8.8).
Remarkably, in comparing the wat er contents of Na
+
-oxidised, Na
+
-reduced,
TMPA
+
-oxidised, and TMPA
+
-reduced smectites, all at the same applied swelling
pressure, they found that the TMPA
+
-reduced sample held the most water! How can
this be, since TMPA
+
is largely a hydrophobic cation (notice that it depresses the
water content when the clay mineral is in the oxidised state)? A plausible explanation
8.3. Surface Interactions with Water 439