Page 238 - Adsorbents fundamentals and applications

PURIFICATION BY π-COMPLEXATION 223

Therefore, the π-complexation sorbents are very promising candidates for
some SMB separations. These separations include: olefins/paraffins (Olex); sep-
aration of hydroxyparaffinic dicarboxylic acids from olefinic dicarboxylic acids;
removal of thiophene, pyridine, and phenol from naphtha; separation of unsat-
urated fatty acid methyl esters from saturated fatty acid methyl esters; and
separation of saturated fatty acids from unsaturated fatty acid. Supported CuCl
+
(such as CuCl/γ -Al 2 O 3 )and Ag (such as AgNO /SiO 2 ) should be good sor-
3
bents for all of these separations, except for the purification of naphtha. For
+
+
the latter application, Cu -zeolite and Ag -zeolites (such as CuY and AgY) are
promising sorbents, as will be discussed in Chapter 10 (under Desulfurization
of Gasoline).
8.5. PURIFICATION BY π-COMPLEXATION

A major difference exists between bulk separation and purification for sorbent
design and selection. For bulk separation, isotherm linearity (hence high-working
capacity) is needed. A steep isotherm (or high Henry’s constant) is needed for
purification. Thus, for π-complexation sorbents, supported salts with monolayer
of full-surface coverage are desirable for bulk separation, whereas ion exchanged
zeolites are suitable for purification.
The extent of π-complexation between the sorbate and sorbent depends, for
a given sorbent, on the density of the π-electrons in the sorbate molecule. Thus,
very strong bonds can be formed with molecules with more than two double
bonds (e.g., dienes), triple bonds, and polynuclear aromatics. At the same time,
for a given sorbate, the sorbent can be tailored to yield a desired bond strength,
by choosing the appropriate cation.
For purification, particularly ultrapurification, a strong adsorption bond is
needed. This means a high Henry’s Law constant is needed. The product purity
in the fluid effluent from a fixed bed adsorber depends on the Henry’s constant,
and the purity can be predicted with a mathematical model (Yang, 1987). For
ultrapurification, e.g., when impurity levels of parts per billion (ppb) or parts per
trillion (ppt) are required, no predictive models exist. This is the case for the
removal of dioxins from effluent in incinerators (Yang et al., 1999).
For the reasons above, the π-complexation sorbents hold a tremendous poten-
tial for future applications in purification, some of which will be included for
discussion. The removal of dienes from olefins by AgY and CuY has already
been demonstrated and applied in the field (Padin et al., 2001). Other promising
applications include:
• desulfurization of gasoline and diesel fuels
• removal of aromatics
• removal of CO from H 2 for fuel cell applications
• VOC removal
• dioxin removal
• removal of acetylene (by Ni 2+ salts)