Layered beds adsorption

For the fluidized bed process the bed expansion as a consequence of an increase in linear flow rate has to be considered. In a simplified picture diffusive transport takes place in a boundary layer around the matrix particle which is frequently renewed, this frequency being dependent on velocity and voidage, as long as convective effects, e.g. the movement of particles are neglected. Rowe [74] has included these considerations into his correlation for kf in fluidized beds, which is applicable for a wide range of Reynolds numbers, including the laminar flow regime where fluidized bed adsorption of proteins takes place (Eq. 19). The exponent m is set to 1 for a liquid fluidized bed, a represents the proportionality factor in the correlation for packed beds (Eq. 18) and is assumed as 1.45. 

The balances of mass of the chemical species i and the terms for the adsorption kinetics (mass transfer, pore diffusion) are listed in Table 9.5-1 for the three systems with Cj as the concentration in the fluid phase and Xj as the mass loading of the adsorbent. J3 denotes the mass transfer coefficient of a pellet and sj, is its internal porosity. The tortuosity factor will be explained later. The derivation of equations describing instationary diffusion in spheres has already been presented in Sect. 4.3.3. With respect to diffusion in macropores it is important to consider that diffusion can take place in the fluid as well as in the adsorbate phase. In Table 9.5-1 special initial and boimdaty conditions valid for a completely unloaded bed (adsorption) or totally loaded bed (desorption) are given. In this section only the model valid for a thin layer in a fixed bed with the thickness dz and the volmne / dz will be derived, see Fig. 9.5-2. 

Because of the wide range of adsorptive properties of the gas molecules in the feed, it was recognized from the early development that more than one sorbent was needed for the separation. Hence, layered beds were used from the beginning. Typically, the first layer (at the feed end) is activated carbon, which is followed by a zeolite (e.g., 5A). The reasoning for such layering becomes obvious from the equilibrium isotherms, shown in Figures 10.21 and 10.22. The most strongly adsorbed components are adsorbed in the activated carbon bed, while the other components are separated in the zeolite bed. 

Numerous studies have been undertaken on the use of layered beds consisting of different sorbents for cyclic adsorption/ion exchange (Klein and Ver-meulen, 1975 Frey, 1983 Wankat and Tondeur, 1985 Chlendi and Tondeur, 1995 Watson et al., 1996 Pigorini and LeVan, 1997). For hydrogen purification using layered activated carbon and zeolite, Chlendi and Tondeur (1995) used the... 

In the adsorption cycle, the wet inlet gas flows downward through the tower. The adsorbable components are adsorbed at rates dependent on their chemical nature, the size of their molecules, and the size of the pores. The water molecules are adsorbed first in the top layers of the desiccant bed. Dry hydrocarbon gases are adsorbed throughout the bed, As the upper layers of desiccant become saturated with water, the water in the wet gas stream begins displacing the previously adsorbed hydrocarbons in the lower desiccant layers. Liquid hydrocarbons will also be absorbed and will fill pore spaces that would otherwise be available for water molecules. 

There are various methods for the determination of the surface area of solids based on the adsorption of a mono-, or polymolecular layer on the surface of the solid. These methods do not measure the particle diameter or projected area as such, but measure the available surface per gram or milliliter of powder. The surface measured is usually greater than that determined by permeability methods as the latter are effectively concerned with the fluid taking the path of least resistance thru the bed, whereas the adsorbate will penetrate thru the whole of the bed as well as pores in the powder particles. These methods appear to be more accurate than surface areas calculated from weight averages or number averages of particle size because cracks, pores, and capillaries of the particles are included and are independent of particle shape and size... 

The Xj is a relative population of adsorption site of type i in the sample and cmax is the Cu+ ions concentration in the sample of the catalyst related to its volume V. F is the rate of flow of the carrier gas, e is a porosity of the layer of the catalyst bed. p is the rate of temperature change. The populations of the Cu+ site types and both desorption energies and desorption entropies for all Cu+ site types were optimized to obtain the best fit with the experimental data. All three experimental Cu-K-FER TPD curves were fitted at once together with all Cu-Na-FER previously measured TPD curves constraining the parameters AHads i and ASads,i to be the same for all samples. 

Although PSA is a batchwise process, by using multiple beds in a sequential manner the overall process is operated in a continuous fashion. Each bed may contain layers of different adsorbent materials selective for specific contaminants in the hydrogen gas stream to be purified. Each bed undergoes a sequence of four basic steps in a PSA cycle adsorption, depressurization, purge at low pressure, and repressurization. This sequence of cyclic operations for each bed is shown schematically for a four-bed PSA process in Figure 8.4 (Yang, 1987 Cassidy, 1980 Miller and Stocker, 1999). 

One of the drawbacks of this CAVERN device is the occurrence of a nonuniform distribution of reactant on catalysts because adsorption occurs on a deep bed of catalyst packed in a MAS rotor. To overcome this problem, we developed several shallow-bed CAVERN devices (95), and Fig. 10 shows a version of one such design. A thin layer of catalyst is supported on a glass trapdoor, and the device is evacuated. A furnace is clamped in place so that the catalyst can be activated if necessary. The catalyst is cooled with a cryogen bath, and a controlled amount of adsorbate is introduced from the vacuum line. The trapdoor is raised, the loaded catalyst falls into the MAS rotor, and the seal is driven into place. Finally the cold, sealed rotor is manually transferred into the cold MAS probe. The added advantages of the shallow-bed CAVERN is that all manipulations can be carried out without using a glovebox in any step. 

One simple way to analyze the performance of a fixed-bed adsorber is to prepare a breakthrough curve (Figure 10.12) by measuring the solute concentration of the effluent as a function of time. As the solution enters the column, most of the solute will be adsorbed in the uppermost layer of solid. The adsorption front will move downward as the adsorption progresses. The solute concentration of the effluent will be virtually free of solute until the adsorption front reaches the bottom of the bed, and then the concentration will start to rise sharply. At this point (tb in Figure 10.12), known as the break point, the whole adsorbent is saturated... 

The activity coefficients were determined after the catalyst had been calcined for three days hi - 9.5151 10u 1/h, Ed = 190431.5 kJ/kmol, s0 = L365. The kinetic constants for the upper and lower layers of the catalytic bed are given in Table 4. The values of the activation energy for the two layers are identical. The adsorption equilibrium constants are the same as those given by Skrzypek at al [13] in their Table 4. 

Abstract. Activated carbon Norit R 08 Extra, and molecular sieve type 4A, were investigated using dynamic (tert-butylbenzene (TBB), cyclohexane (CHX) and water vapour) adsorption methods. The TBB, CHX and water breakthrough plots for fixed activated carbon - molecular sieve beds were analyzed. It was found that the type of bed composition with mechanically mixed activated carbon with molecular sieve, or separated activated carbon and molecular sieve layers, affects the dynamic adsorption characteristics. 

---