Page 60 - Adsorbents fundamentals and applications

SIMPLE CRITERIA FOR SORBENT SELECTION 45

Table 3.3. PSA simulation operating conditions, performance results, and parameter
+
S for the comparison of LiX (Al/Si = 1, 100% Li -exchange) vs. NaX adsorbents for air
separation (P H = 1.0 atm, average product purity = 95.5%, average product throughput =
2.7 × 10 −2 kg O 2 /h/kg sorbent)
+
(A) LiX or LiLSX (Si/Al = 1, 100% Li ) adsorbent
Pressure Ratio Purge-to-Feed P CD O 2 Product Sorbent
Ratio (atm) Recovery Selection
P H /P L
P/F (%) Parameter, S
2 0.44 0.80 53.5 201.7
3 0.23 0.70 61.9 220.9
4 0.13 0.70 63.3 232.0
5 0.09 0.70 64.1 239.3
6 0.06 0.65 65.3 244.4
7 0.04 0.65 65.3 248.1
8 0.03 0.65 64.5 251.0
10 0.02 0.65 64.7 255.2
(B) NaX (13X) adsorbent
Pressure Ratio Purge-to-Feed P CD O 2 Product Sorbent
P H /P L Ratio (atm) Recovery Selection
P/F (%) Parameter, S
2 0.80 0.85 20.0 110.40
3 0.49 0.80 42.3 118.29
4 0.46 0.70 53.3 122.82
5 0.35 0.70 53.4 125.77
6 0.27 0.70 53.2 127.83
7 0.21 0.70 52.8 129.36
8 0.18 0.67 52.8 130.54
10 0.14 0.67 53.1 132.24
From Rege and Yang, 2001.

Although the two sorbents have a subtle difference in performance, the better
one involving the LiAgX sorbent is reflected accurately by the parameter. In both
cases S was higher for LiAgX compared with LiLSX under the said operating
pressures. Thus the sorbent selection parameter is shown to be sensitive to even
slight differences in the adsorption isotherms.
The S parameter has been defined for Langmuir systems, that is, the isotherms
for both components are Langmuirian. For non-Langmuirian systems, α 12 proba-
bly should be replaced by α 2 /α des , as given by the AFM, because this ratio was
ads
empirically determined. It seems that the α during the adsorption step is more
important than during the desorption step in determining the PSA performance.