Page 53 - Adsorbents - fundamentals and applications

38 SORBENT SELECTION: CRITERIA

et al., 1996; Hufton et al., 1999; 2001) for the feasibility of hydrogen produc-
tion from the water gas shift reaction. Their sorbent was hydrotalcite for the
adsorption of CO 2 , and a noble metal as the catalyst at a bed temperature near
◦
490 C. The enhancement of catalytic reactions by combining with PSA has been
analyzed and discussed in detail by Alpay and co-workers (1993; Chatsiriwech
et al., 1994; Cheng et al., 1998; Yongsunthon and Alpay, 1998, 2000). They also
studied the PSA reactors for steam reforming of methane by using hydrotalcite
CO 2 sorbent and supported Ni catalyst (Ding and Alpay, 2000) and l-butene
dehydrogenation by using CrO 2 -alumina catalyst and a zeolite K-Y adsorbent
(Sheikh et al., 2001). Chou et al. (1998) performed a simulation of the PSA
reactors for CO oxidation (i.e., the system studied by Vaporciyan and Kadlec,
1989) and cyclohexane dehydrogenation (Chou et al., 1998). The series reac-
tion A → B → C was studied with the PSA reactor by removing A by Kodde
et al. (2000).
The PSA reactor clearly holds potential for equilibrium-limited reactions, such
as water–gas shift and dehydrogenation reactions. However, the limiting factor
remains to be the sorbent. The PSA reactor must be operated at a relatively high
temperature in order for the catalyzed reaction to proceed at a realistic rate. High
temperature is not favorable for sorption (or sorption reaction). The challenge
for this problem is to find the proper sorbent that could react and sorb one of
the products at a high rate and with a high capacity. At the same time, the
sorbent must be readily regenerable. For H 2 production by water gas shift, a
better sorbent than hydrotalcite is needed for CO 2 (forming carbonate), and the
◦
reaction needs to be run at a temperature higher than 500 C for reasonable rates.
Other alkaline earth oxides and perovskites may hold promise for this application.
Sorbent integrity upon cyclic sorption/desorption (or decomposition) can also be
a severe limitation.

Analytical Solution for Skarstrom Cycle. It is desirable to have a simple,
analytical solution for PSA that can be used for estimation purposes, e.g., for
sizing the beds and comparing different sorbents. As Knaebel quoted Chilton,
“The simpler things become in a piece of research or development, the closer
one has come to the truth.” (Ruthven et al., 1994). The following is the solution
by Chan et al. (1981) which has proved to be quite useful. The PSA cycle is a
simple two-step Skarstrom cycle with linear isotherms.
The assumptions for the model are

1. Linear isotherms are followed by both components A and B. The strong
adsorptive, A, is at a trace level. The two isotherms are non-interfering.
2. The cycle is isothermal.
3. The interstitial flow velocity, u, is constant during the adsorption and
purge steps.
4. Heat and mass transfer are instantaneous.
5. Plug flow is assumed, with no axial or radial dispersion.