Stiffening pores

Sakohara et al. [1989] used this method to prepare essentially "leak-prooP silica-alumina membranes modified with PVA or CA. The silica-alumina membranes were made by treating sol-gel alumina membranes with sodium silicate. The treated membranes were then soaked in a dilute solution of PVA (in water) or CA (in acetone) at room temperature with one side of the membrane under vacuum to ensure contact of PVA or CA with the pinholes. In the case of CA, exposure to hot water for a short time is necessary to stiffen CA which resides in some large pores. The above procedure involving the organic sealants may be repeated a number of times to achieve an acceptably low level of solvent leakage through the membrane. 

Figure 10. Illustration of drying process (24). Capillary tension develops in liquid as it stretches to prevent exposure of the solid phase by evaporation, and the network is drawn back into liquid (a). The network is initially so compliant that little stress is needed to keep it submerged, so the tension in the liquid is low, and the radius of the meniscus (rc) is large (b). As the network stiffens, the tension rises as rc decreases. At the critical point, the radius of the meniscus becomes equal to the pore radius the constant rate period ends and the liquid recedes into the gel (c). (Reproduced with permission from reference...

Weatherwax RC., Caufield DF (1978) The pore stmcture of papers wet stiffened by fimnaldehyde crosslinking. Journal of colloid and int ace science 67 498 505. 

Let us consider a wet gel in which some liquid suddenly evaporates. As outlined earlier, the liquid in the pores stretches to cover the dry region and a tension develops in the liquid. The tension is balanced by compressive stresses on the solid phase of the gel. Since the network is compliant, the compressive forces cause it to contract into the liquid, and the liquid surface remains at the exterior surface of the gel (Fig. 5.18b). In a polymeric gel, it does not take much force to submerge the solid phase, so that initially the capillary tension of the liquid is low and the radius of the meniscus is large. As drying proceeds, the network becomes stiffer because new bonds are forming (e.g., by condensation reactions) and the porosity is decreasing. The meniscus also deepens (i.e., the radius decreases), and the tension in the liquid increases [Eq. (5.39)]. When the radius of the meniscus becomes equal to the pore radius in the gel, the liquid exerts the maximum possible stress [Eq. (5.41)]. This point marks the end of the CRP beyond this the tension in the liquid cannot overcome the further stiffening of the network. The liquid meniscus recedes into the pores, and this marks the start of the FRP (Fig. 5.18c). Thus, the characteristic features of the CRP are... 

Another approach is to strengthen the gel so that it is better able to withstand the drying stresses. For example, aging the gel in the pore liquid at slightly elevated temperatures stiffens the gel network and also reduces the amount of shrinkage in the drying stage. 

When the gel is aging, new bonds can also be formed between neighboring branches of the network, which are brought into contact by thermal fluctuations. As a consequence, the gel not only stiffens, but also shrinks linearly by a few percent and expels some of the pore liquid this is called syneresis (Scherer, 1999). The described synthesis and postprocessing of silica gels is performed at (or slightly above) room temperature. 

In the above expressions for gel stress, the liquid pressure level does not appear explicitly. Nevertheless, capillary pressure plays a crucial role because it decides to what extent the material will shrink, and thereby determines how much the gel stiffens and by how much its permeability is reduced. Consequently, lowering the capillary pressure - by increasing pore size, reducing surface tension or increasing contact angle - is beneficial for product quality because it reduces the maximum tensile stress. 

It should, however, be kept in mind that the model crack criterion contains no contribution from the drying-out of the gel surface - be it in terms of an averaged receding liquid/gas boundary or, at the microscale, as information about the emptying of individual pores. In fact, the simulation predicts cracks because, during stiffening of the gel, stresses increase more than the strength (see Fig. 5.42). 

Figures 11J5 and 11.16 illustrate two designs of Leclanche battery. The first shows the traditional cylindrical design. The negative zinc electrode is a zinc lining to the metal can which is amalgamated with mercury to minimize hydrogen gas formation by reaction of the metal with water the separator is a paper stiffened with cellulose or starch placed adjacent to the zinc can. The positive current collector is a carbon rod at the centre of the can, while most of the volume is taken up by the positive paste. This is a mixture of powdered manganese dioxide ainmotmtm chloride and acetylene black (carbon) to increase the conductivity the pores are filled with an aqueous electrolyte (NH Cl + ZnCl ) gelled by addition of starch. The can is totally scaled.

In this case, the stress does not depend on the thickness of the plate, because the permeability is so low that the pressure drops to zero between the surface and the midplane. During drying the network stiffens, the effective viscosity of the pore liquid rises (because a larger proportion of it is near a solid surface), and the permeability decreases (because of the decrease in pore size and porosity), and all these factors contribute to a continual increase in. stress. However, unless the gel is quite large, it seems unlikely that a will be large enough for Eq. 30 to apply. 

The same sort of experiments were performed [15] using a two-step base-catalyzed gel called B2 [14] (made from TEOS with r = 3.7, pH 7.9). When soaked in the pore liquid (89 vol% ethanol and 11 % water) the shear modulus increased for as much as one year, as shown in Fig. 26 the rate of stiffening was greater at 35°C than at room temperature. Other samples were soaked in various ethanol-water solutions or allowed to dry partially before measurement. Samples aged in pure ethanol or in ethanol-acetone solutions showed neither shrinkage nor increase in modulus over a period of... 

---