Page 317 - Adsorbents fundamentals and applications

302 SORBENTS FOR APPLICATIONS

other materials, O 2 was strongly bound at very low pressures. However, as with
all of the other immobilized samples, it does not immediately attain the full
capacity, but instead has a positive slope through 1 atm. The N 2 adsorption
isotherm was also measured for this Co(fluomine)-IXR material and is shown in
Figure 10.19. The theoretical capacity for this material is about 1.2 mmol/g for a
2+
2+
1: 1 (O 2 :Co ) and 0.6 mmol/g for a 1 : 2 (O 2 :Co ) adduct. These values assume
100% ion-exchange of Co 2+ with the ion-exchange resin (with a cation exchange
capacity of 4.7 meq/g) and subsequent conversion to Co(fluomine). The actual
◦
capacity of this material, at 0.03 mmol/g at 25 C and 1 atm, is considerably
lower than the theoretical amount.
The immobilized samples were found to be very stable when exposed to
atmospheric moisture. There was no noticeable decrease in the binding capacity
after exposure to atmospheric oxygen and moisture. In order to test the stabil-
ity to repeated oxygenation and deoxygenation, the samples were subjected to
an aggressive stability test. Each of the samples was subjected to oxygenation
and deoxygenation cycles in a thermogravimetric analyzer (TGA). The samples
◦
were first activated by heating to 120 C in a pure Ar atmosphere. The samples
◦
were then alternatively exposed to 1 atm O 2 and 1 atm Ar at 60 C for 15 min
each. After 90 cycles, the samples were removed from the TGA and reactivated
◦
at 120 C in a vacuum. The oxygen adsorption and desorption isotherms were
then measured, and the sample was then returned to the TGA for continued
cycling. The results are summarized in Figure 10.20. In each free and immo-
bilized Co(fluomine) synthesis, piperidine was added as the last step. This was
done in order to follow previously reported procedures (Bailes and Calvin, 1947),
although samples prepared with and without this step did not show any differ-
ence (Hutson and Yang, 2000b). The results of the accelerated test for stability
showed that the immobilized samples still lacked stability. However, they seemed
to stabilize at about 40% of their initial capacities.
Attention has been given to the possibility of air separation based on magnetic
properties. Oxygen is paramagnetic, while nitrogen and argon are not. Hence oxy-
gen might be repelled magnetically from a diamagnetic superconductor, that is,
the Meissner effect. Makarshin et al. (1997) reported that the BET surface area of
a superconducting material measured from O 2 adsorption was about 6 times lower
than that from N 2 . Gordon and Cussler (1999a) measured the isotherms of N 2 and
O 2 on two superconducting sorbents and compared them with those on graphite
and SiO 2 . No noticeable difference in the isotherms was observed. They also
examined the possibility of air separation by using a membrane fabricated from
a superconducting material (Gordon and Cussler, 1999b). The results of these
experiments were complicated by the condensation of both O 2 and moisture,
which explained earlier results as well. The conclusion was that the possibility
of air separation with superconducting materials was not promising. A related
study was performed by Ozeki et al. (2000), who measured the adsorption of
O 2 on sorbents with external magnetic fields (from 1 T to 6 T). Like the exper-
iment of Gordon and Cussler, the isotherms were measured at 77 K, below the
onset of superconductivity of YBa 2 Cu 3 O 7-x . Interestingly, with 1 : 1 mixture of