Preferential adsorption, multicomponent

Accounting for size differences can also be realized in terms of distribution functions, assuming certain interaction energies. Simply because of size differences between molecules preferential adsorption will take place, i.e. fractionation occurs near a phase boundary. In theories where molecular geometries are not constrained by a lattice, this distribution function is virtually determined by the repulsive part of the interaction. An example of this kind has been provided by Chan et al. who considered binary mixtures of adhesive hard spheres in the Percus-Yevick approximation. The theory incorporates a definition of the Gibbs dividing plane in terms of distribution functions. A more formal thermodynamic description for multicomponent mixtures has been given by Schlby and Ruckenstein ). 

In most situations the experimental system is more complicated than one (homodisperse) polymer adsorbing from a single solvent. In multicomponent systems preferential adsorption always plays a role. A common example is the adsorption of a polydisperse polymer, where usually long chains adsorb preferentially over short ones, even if the adsorption energy per segment is the same. 

The FHT measured by the Wilhelmy balance method can be effectively used to compare the liquid holding capabilities of different surfaces. The value of FHT depends on the experimental parameters and cannot be used in an absolute sense. In general, the aqueous film stability is obtained when spontaneous wetting occurs on imperturbable surfaces. However, moderately hydrophilic and possibly even some hydrophobic surfaces that are perturbable by water were found to be capable of holding continuous films of water. Multicomponent fluid, such as a dilute solution of protein used as a simulated tear fluid, may yield misleading liquid holding characteristics of surfaces due to preferential adsorption of components on a surface. 

In summary, one can see that separation selectivity for gas and vapor molecules depends on the category of pores (mesopores, supermicropores, and ultramicropores) and on the related transport mechanisms. Either size effect or preferential adsorption effect (irrespective of molecular dimension) is involved in selective separation of multicomponent mixtures. The membrane separation selectivity for two gases is usually expressed either as the ratio between the two pure gas permeation fluxes (ideal selectivity) or between each gas permeation flux measured from the mixture of the two gases (real selectivity). More detailed information on gas and vapor transport in porous ceramic membranes can be found in Ref. [24]. 

In multicomponent systems (i.e., more than two components in the solution), there is in addition preferential binding (or preferential adsorption) of the solvent... 

Most real protein-containing fluids of interest, such as blood, are multicomponent systems so that the influence of one protein on another may become important. With respect to adsorption, the main consideration is whether adsorbed amounts are in proportion to solution concentration or whether surfaces "select one protein in preference to another. Presumably if proteins act independently of each other one should be able to predict adsorbed amounts in mixtures from relative affinities derived from single protein studies. Although there have been no systematic attempts to make such predictions it seems likely that they would fail. In general it has been found that preferential or selective adsorption occurs so that certain proteins may be enriched in the surface relative to the solution and vice versa. There have as yet been no attempts to determine the properties of protein-surface systems that govern the relative surface affinity of different proteins. More will be said on this topic when adsorption from plasma is discussed. 

In multicomponent systems (e.g., surfactant solutions), surface tension gradients usually are due to adsorption-related phenomena or, where possible, to different rates of evaporation from the system (although simple temperature variations can also be important). If the system contains two liquid components of differing volatility, the more volatile liquid may evaporate more quickly from the LV interface, resulting in localized compositional—and therefore surface tension—differences. It is also commonly found that when two or more components are present, one will be preferentially adsorbed at the LV interface and lower ctlv of the system. If a surface-active component... 

Multicomponent systems may also involve the selective adsorption of one component at the SL interface. Since the component that lowers the interfacial tension will be preferentially adsorbed, the rate of the adsorption process can affect the local tension and the contact angle. In many systems, the rate of adsorption at the solid surface is found to be quite slow compared to the rate of movement of the SLV contact line. As a result, the system does not have time for the various interfacial tensions to achieve their equilibrium values. Most surfactants, for example, require several seconds to attain adsorption equihbrium at a LV interface, and longer times at the SL interface. Therefore, if the hquid is flowing across fresh solid surface, or over any surface at a rate faster than the SL adsorption rate, the effective values of olv and osl (and therefore 6) will not be the equilibrium values one might obtain from more static measurements. More will be said about dynamic contact angles in later chapters. 

The described properties of the potential theory of adsorption make it preferential to the theory of localized adsorption, at least, for complex thermodynamic conditions. On the other hand, the potential theory incorporates specific physical information about the phenomenon of adsorption, which is absent in the general thermodynamic approach. A major disadvantage of the multicomponent potential theory is the computational difficulties arising in the calculation of the multicomponent segregation in the external field. This is the reason why the development of this theory has been delayed until our recent works [92-94]. The development of modem methods of computational thermodynamics [68] has made it possible to perform these calculations quite efficiently. 

---