Page 55 - Adsorbents - fundamentals and applications

40 SORBENT SELECTION: CRITERIA

Starting from saturated beds as the initial condition, Chan et al. (1981) derived
solutions for product purity and recovery:
Product Purity:

The mole fraction in the nth cycle is
Y n = Y f (P H /P L ) 2n(β−1) (3.54)

where Y f is the mole fraction of A in the feed, and β is given by:

β A
β = < 1 (3.55)
β B
Product Recovery (ρ):

(P H /P L ) 1−β − 1
ρ = (3.56)
(P H /P L ) 1−β (βL/Z L )(P H − P L )/P L
where L is the length of the beds and Z L is given by Eq. 3.53.
The bed size is given by the bed length and hence the sorbent productivity
can be calculated from the feed velocity.

3.3. SIMPLE CRITERIA FOR SORBENT SELECTION
Detailed mathematical models are needed for accurately assessing a given sorbent
for either PSA or TSA. Many models are available in the literature (Yang, 1987;
Ritter and Al-Muhtaseb, 2001). A number of commercial computing packages
are already available, such as ADSIM (by Aspen) and SiMoBe (by Prosim).
During the process development stage, several adsorbents may be identified
as suitable materials to effect a viable separation for a particular application.
The preliminary viability can be judged by a visual comparison of the isotherms
and/or the uptake rates of the pure components of the gas mixture. However,
a cursory visual inspection can be misleading and does not quantify the merit
or the efficiency of the sorbent for the separation under consideration. For a
detailed comparison of the adsorbent performance, it is necessary to perform lab
scale PSA experiments by using a suitable cycle, or computer simulations with a
proven mathematical model, as mentioned. It is obvious that these steps are time-
consuming. A quick and easily calculable parameter would be highly desirable
in selecting the optimum sorbent for a particular gas-separation application. A
simple selection parameter for sorbent selection for PSA has been given by Rege
and Yang (2001a and b), and will be given here.
A simple method for determining sorbent selectivity was proposed by Knaebel
(1995), which involved taking the ratio of Henry law constants. This rough esti-
mate for selectivity can serve as a crude sorbent selection parameter. A more
precise estimate can be obtained by comparing the isothermal binary working