Page 63 - Adsorbents fundamentals and applications

48 SORBENT SELECTION: CRITERIA

This parameter is incomplete because it does not include the actual amount
adsorbed during the adsorption step. The sorbent productivity is directly depen-
dent upon the amount adsorbed. Thus a complete parameter should include this
amount, as

2
K 1 D 1 6K P 1 D 1 D 1 t
1
S k = q t = (3.67)
2 2
K 2 D 2 π K 2 D 2 r
c
Here S is the sorbent selection parameter, component “1” is the fast diffusing
component, t is the step time for adsorption, and P is the partial pressure in the
feed mixture. For a given feed mixture and step time t, the parameter may be
simplified as follows when comparing different sorbents:
K 1 2 D 1
S k = √ (3.68)
K 2 D 2

2
where D may be replaced by D/r .
c
Neither of these parameters has been tested. The main commercial kinetic sep-
aration is nitrogen production from air by using molecular sieve carbon. The cycle
used is a simple Skarstrom cycle. This parameter may be tested for this system.
As discussed, for the full assessment of a sorbent, PSA simulation is needed.
Kinetic PSA simulations with various degrees of simplifications are available. For
instance, Schork et al. (1993) reported a kinetic PSA model with the assump-
tion that the solid-phase concentration profiles along the bed are linear, and the
computation can be substantially reduced. Their simplified model was used to
evaluate carbon molecular sieves for N 2 production from air.

NOTATION

a partial molar area
A surface area of sorbent
b Langmuir constant
B Langmuir constant
c concentration
C dispersion constant; average number of sorbate molecules per cage
in zeolite
D diffusivity
D s surface diffusivity
D so surface diffusivity at zero coverage
E interaction energy
f fugacity; partition function
f , fugacity at saturation
F field strength
G Gibbs free energy
h Planck constant