Relationship of capillary pressures to pore radii and liquid-water contents
Defay and Prigogine (1966) have derived a form of the Laplace equation for
freezing liquids in saturated porous media:
2γ ls
pc =
(21)
r ls
where rls is the radius of a concave liquid/solid interface in a pore. Brun et al.
(1977) found that rls is equal to the pore radius less the few molecular diameters
corresponding to the liquid that wets the pore solid surfaces, which remains
unfrozen to temperatures well below the bulk melting point. Experimentally mea-
sured freezing curves of porous solids are consistent with a general rela-
tionship between liquid/solid capillary pressure and the specific volume
of unfrozen liquid. For a given capillary pressure corresponding to a particular
pore radius, it is assumed generally that the contents of all narrower pores are
unfrozen and the contents of all wider pores are solidified. This capillary
pressureunfrozen liquid relationship is physical and should be unaffected directly
by the chemical composition of the phases. For a given porous material, there-
fore, it is expected that this relationship will be general and unaffected by
phase composition. We tested this expectation with experiments described in the
following section.
MATERIALS AND METHODS
Sample preparation
Kaolinite clay
Approximately 40 g of clay were washed five times with 200 mL of 1-mol kg1
NaCl aqueous solution to saturate the cation exchange complex with Na+. In each
washing, the claysolution suspension was stirred for about 515 minutes with a
vortex stirrer. The solids were separated with a centrifuge.
Following the last washing with 1-mol kg1 NaCl solution, the suspension was
separated into three aliquots and individually washed with 0.1-, 0.01-, and 0.001-
mol kg1 NaCl solutions, respectively. These washings were followed by centrifu-
gation; this process was repeated four times. The individual samples were then
sealed in test tubes with rubber stoppers to prevent moisture changes.
Montmorillonite clay
Sodium-saturated montmorillonite that had been prepared previously for
other experiments and then stored was selected for analysis. Three 4-g
samples of this clay were placed in beakers to which 200-mL solutions of
either 0.1-, 0.01-, or 0.001-mol kg1 NaCl were added. The clays were then sepa-
rated from the solutions and resuspended five times with further washings of 200-
mL solutions of 0.1-, 0.01-, and 0.001-mol kg1 NaCl. The 0.1-mol kg1 suspension
was centrifuged at high speed for approximately 6 hours. The liquids in the 0.01-
and 0.001-mol kg1 suspensions were separated from the clay by pressure-mem-
brane extraction. Following the above treatments, the samples were placed in sealed
test tubes.
Quartz sand
A fine-textured sand was divided into three 50-g subsamples that were washed
four times with either 0.1-, 0.01-, or 0.001-mol kg1 NaCl solution. Each saturated
sand sample was sealed in test tubes.
5