Adsorptive separations are generally classified as either equilibrium-or kinetic-based depending on the importance of thermodynamic adsorption affinity relative to mass transfer limitations (Yang, 1987). When mass transfer resistances are small, local equilibrium between gas and solid may be assumed and adsorption affinity dominates. Kinetic separations assume that diffusional resistances are important resulting in the formation of concentration gradients within the adsorbent. The linear driving force (LDF) model makes the simplifying assumption that the intraparticle concentration profile is parabolic in shape (Liaw, 1979;Doong and Yang, 1986). Recently, Doong and Yang (1986) developed a bidispere pore diffusion model, in which the parabolic profile assumption has been employed to describe both macropore and micropore concentration profiles. The parabolic profile assumption is, however, valid only for long times after step changes, D,t/r? > 0.1.
The kinetic separation of air (mainly for producing nitrogen) using molecular sieve carbon has been investigated by Jiingten (1977) and Hassan et al. (1986). Recently, the separation of nitrogen from air using 4A zeolite molecular sieve has also been described as a kinetic separation (Pan et al., 1988;Shin and Knaebel, 1987) because of the small micropore diameter of this adsorbent (4A) which approaches the molecular diameter of the adsorbates. Because of the small pore size, nitrogen has a significantly lower micropore diffusivity than does oxygen in these adsorbents. Kinetic separations attempt to take advantage of this disparity in diffusion rate to offset the higher equilibrium loading of nitrogen.
Concentration breakthrough curves in fixed-bed, multicomponent adsorption systems are used often to test the validity of adsorption models. A common characteristic of such breakthrough curves is the phenomenon of rollup wherein a lighter, weakly adsorbed component is displaced by a heavier, more strongly adsorbed component (Ruthven